GeMMA Activity Report 2016–2022 Tamara Golob GeMMA Activity Report 2016–2022 Tamara Golob March 2023 eMA Title GeMMA Activity Report 2016–2022 Author Tamara Golob (University of Maribor, Faculty of Electrical Engineering and Computer Science) Language editing Marko Bizjak Laboratory members Employees of University of Maribor, Faculty of Electrical Engineering and Computer Science, Laboratory for Geospatial Modelling, Multimedia and Artificial Intelligence listed in chapter Members. Technical editors Tamara Golob (University of Maribor, Faculty of Electrical Engineering and Computer Science) Jan Perša (University of Maribor, University Press) Cover designer Tamara Golob (University of Maribor, Faculty of Electrical Engineering and Computer Science) Cover graphics Information Annual Report, https://elements.envato.com/information-annual-report-BGJD5KA, Author Username: aqrstudio, License Code: YKRM4X3CGT. Cyber circuit future technology concept background Pro Photo, https://www.vecteezy.com/photo/7882966-cyber-circuit-future-technology-concept-background, Author: Titima Ongkantong, License Code: 7882966. Graphic material University of Maribor, Faculty of Electrical Engineering and Computer Science, Laboratory for Geospatial Modelling, Multimedia and Artificial Intelligence listed & Golob, 2023 Photographer Bogdan Dugonik (photographs pp. 4-21) Published by University of Maribor University Press Slomškov trg 15, 2000 Maribor, Slovenia https://press.um.si, zalozba@um.si Issued by University of Maribor Faculty of Electrical Engineering and Computer Science Koroška cesta 46, 2000 Maribor, Slovenia https://um.feri.si, feri@um.si Edition 1st Publication type E-book Published at Maribor, Slovenia, March 2023 Available at https://press.um.si/index.php/ump/catalog/book/764 © University of Maribor, University Press / Univerza v Mariboru, Univerzitetna založba Text © Golob, 2023 This book is published under a Creative Commons 4.0 International licence (CC BY 4.0). This license lets others remix, tweak, and build upon your work even for commercial purposes, as long as they credit you and license their new creations under the identical terms. This license is often compared to “copyleft” free and open source software licenses. Any third-party material in this book is published under the book’s Creative Commons licence unless indicated otherwise in the credit line to the material. If you would like to reuse any third-party material not covered by the book’s Creative Commons licence, you will need to obtain permission directly from the copyright holder. https://creativecommons.org/licenses/by/4.0/ CIP - Kataložni zapis o publikaciji Univerzitetna knjižnica Maribor 001.892:004.8(082)(0.034.2) GOLOB, Tamara, računaln. GeMMa Activity Report [Elektronski vir] : 2016-2022 : March 2023 / [author] Tamara Golob ; [photographer Bogdan Dugonik]. - 1st ed. - Maribor : University of Maribor, University Press, 2023 Način dostopa (URL): https://press.um.si/index.php/ump/catalog/book/764 ISBN 978-961-286-713-3 doi: 10.18690/um.feri.2.2023 COBISS.SI-ID 143516675 ISBN 978-961-286-713-3 (pdf) 978-961-286-714-0 (softback) DOI https://doi.org/10.18690/um.feri.2.2023 Price Free copy For publisher Prof. Dr. Zdravko Kačič, rector of University of Maribor Attribution Golob. T. (2023). GeMMA Activity Report 2016–2022. University of Maribor, University Press. doi: 10.18690/um.feri.2.2023 Table of Contents Foreword 1 Members 4 GeMMA Team 5 Part-time Members 21 Past Members 22 Projects 23 Slovene National Project 24 International Projects 71 Bilateral International Projects 90 Industrial Projects 100 Achievements 123 Publications 124 Prizes, Awards, Honours, Medals 136 Ph.D. Candidates Granted by ARRS and Completed Ph.D.s Supervised in GeMMA 138 Functions and Honours in National and International Associations 140 I Acronyms 2D 2-dimensional 2.5D 2.5-dimensional 3D 3-dimensional ACM Association for Computing Machinery AI Artificial intelligence AIOTI Alliance for the Internet of Things Innovation API Application programming interface AR Augmented reality ARRS Slovenian Research Agency ARSO Slovenian Environment Agency BDVA Big Data Value Association BIM Building Information Modelling B.Sc. Bachelor of Science CGAI Laboratory for Computer Graphics and Artificial Intelligence CH Cultural heritage CNN Convolutional neural networks COMPROMISE Data Compression Paradigm Based on Omitting Self-evident Information DAIRO Data, AI and Robotics DNA Deoxyribonucleic acid d.o.o. Company with limited liability (Sl Družba z omejeno odgovornostjo) EC European Commission ELES Slovenian electricity transmission system operator EMS Energy management system EO Earth observation II ERDF European Regional Development Fund ERK International Electrotechnical and Computer Science Conference ESGO European Society of Gynaecological Oncology ESP European Society of Pathology ESRI Environmental Systems Research Institute ESTRO European Society for Radiotherapy & Oncology ETSI European Telecommunications Standards Institute EU European Union EUROGI European Umbrella Organisation for Geographic Information FERI Faculty of Electrical Engineering and Computer Science FGPA Faculty of Civil Engineering, Transportation Engineering and Architecture FL Federated learning FST Fertility sparing therapy GDP Gross domestic product GeMMA Laboratory for Geospatial Modelling, Multimedia, and Artificial Intelligence GEOSS Global Earth Observation System of Systems GFS GeMMA Fusion Suite GHG Greenhouse gas GIS Geographic Information System GISIG Geographical Information System International Group GPU Graphics processing unit GZ Construction Law (Sl Gradbeni zakon) H2020 Horizon 2020 HPC High-performance computing ICPC International Collegiate Programming Contest ICS Intelligent control systems IEEE Institute of Electrical and Electronics Engineers IMINT Imagery intelligence III INSPIRE Innovation in Science Pursuit for Inspired Research IOT Internet of things IWW Inland water ways KPI Key performance indicator LiDAR Light detection and ranging MGRT Ministry of Economic Development and Technology MIZŠ Ministry of Education, Science and Sport MKGP Ministry of Agriculture, Forestry and Food ML Machine learning MNZ Ministry of the Interior MOP Ministry of the Environment and Spatial Planning MORS Ministry of Defence M.Sc. Master of Science MZI Ministry of Infrastructure OGC Open Geospatial Consortium PLACE Predictive Analytics Based on Location-associated Context Enrichment Ph.D. Doctor of Philosophy PRO Patient reported outcomes R&D Research and development RDP Research and development project REST Representational state transfer RNA Ribonucleic acid RS Republic of Slovenia S4 Slovenian Smart Specialisation Strategy SCI Science Citation Index SERŠ Secondary School of Electrical Engineering and Computer Science Maribor SMA Surveying and Mapping Authority TRL Technology readiness level IV UAV Unmanned Aerial Vehicle UI User interface UL University of Ljubljana UM University of Maribor UMC University Medical Centre UX User experience UWB University of West Bohemia VGI Volunteered geographic information VLC Visual light communication WP Work package WWW World Wide Web ZUreP-2 Spatial Planning Act (Sl Zakon o urejanju okolja) V GeMMA Activity Report 2016–2022 VI GeMMA Activity Report 2016–2022 Foreword The predecessor of the Laboratory for Geospatial Modelling, Multimedia and Artificial intelligence, the Laboratory for Geometric Modelling and Multimedia Algorithms (GeMMA), was established on 1st January 2000. The very early days of the Laboratory, initially having only three members, were characterized with minimal finances and the most elementary equipment (three the most basic personal computers). The initial research focus was on the algorithms of 2D computational geometry and 3D geometric modelling, especially their efficient and stable implementation in various applications including civil and mechanical engineering, medicine, digital cultural heritage, and geographical information systems (GIS). One of the most remarkable achievements was, when two of our algorithms were incorporated into the AutoDesk development framework (namely, polygon triangulation and polygon trapezoidation). Soon we established a strong cooperation with the Slovene GIS company Igea, d.o.o., and participated in the first European Union supported project Virtual Heart of Central Europe. This, however, gave us the possibility to employ full-time researchers and gain a steep growth of the Laboratory. Over the years, GeMMA’s research areas have also expanded, including remote sensing data processing, GIS, scientific visualization, data compression, simulation of green-energy resources, big data processing, predictive analytics, and data mining. In 2016, GeMMA unified with the Laboratory for Computer Graphics and Artificial Intelligence (CGAI). With this reunion (the three original GeMMA members also used to work in CGAI before 2000), GeMMA became the strongest research laboratory at the Institute for Computer Science. Since GeMMA was founded, 25 PhDs have been graduated under the supervision of its members. The majority used to work in GeMMA for some period (or they still do). On this basis, the Laboratory has built a team of experienced project leaders who are gaining new research projects at national, European, and industrial level. As the main research orientations changed, the Laboratory was renamed in 2021. GeMMA has 28 full-time employees (10 PhDs, while half of the rest are PhD students) and two part-time members (2 PhDs) at the end of 2022. The primary aim of this survey is to support the dissemination of the research achievements of GeMMA. It follows the first survey published in 2016, and therefore, the actual book concentrates on the research results since then. The previous book, covering a period of 17 years, presented 35 projects, while the new review of activities over the last 7 years covers as many as 58 projects. This growth is a good cue to introduce the secondary, equally important aim of the book. Namely, we would like to leave the track to our successors to stimulate them for even better research results, to show them, what is possible to achieve in 22 years starting from scratch with the will, hard work, orientation towards the applications, devotion to the research work, and the strong team spirit. Maribor, November 8th, 2022 Borut Žalik Head of GeMMA 1 GeMMA Activity Report 2016–2022 GeMMA in Numbers since 2016  database book gear globe  over over over over 5.5 108.4 13 20 50 150 mil € bn. New members International National, Publications of Revenue from Cups of of our research international, last 6 years projects carried coffee the lab laboratory groups we have bilateral and out in the past members drank since 2016 worked with in industrial 6 years during working the past 6 years projects in the hours in the past 6 years past 6 years Figure 1: GeMMA in Numbers since 2016 (Source: own). 2 GeMMA Activity Report 2016–2022 Figure 2: A look inside our lab with DroneVIS, tree growth simulator and some projects on display. (Foto: Ž. Ivanc) Figure 3: The other side of our lab with hologram and some projects on display. (Foto: Ž. Ivanc) 3 MEMBERS GeMMA Activity Report 2016–2022 GeMMA Team dr. Borut Žalik Borut Žalik graduated in electrical engineering in 1985 at the Technical faculty in Maribor. The same year he was employed at the same Faculty as a technician and a few months later as a teaching assistant. He obtained M.Sc. and Ph.D. in Computer Science from the University of Maribor in 1988 and 1993, respectively. In 1993, he was elected as an Assistant professor of Computer Science and 5 year later he became Associate professor. In 2003, he was elected as a full professor of Computer Science at the Faculty of Electrical Engineering and Computer Science, University of Maribor (UM FERI). He spent half a year at Technical University Graz (Austria) as a Research fellow in 1992. From 2000 to 2002 he has been the Visiting Research Fellow at De Montort University, U. K. He became the Head of the Laboratory for Geometric Modelling and Multimedia Algorithms (GeMMA) in 2000. Laboratory was renamed in Laboratory for Geospatial Modelling, Multimedia and Artificial Intelligence due to the change in the research focus in 2021. He was the Vice-Dean of Research from 2003 to 2011, and since 2011 to 2019 he was a Dean, both at UM FERI. He was a member of the management board of the Slovenian Research Agency in 2011 and 2012. In 2014 he became the member of the European Academy of Sciences and Arts. He was honoured with the designation of the ACM Senior Member in 2020. In 2021, he became the member of the Professional board of the Maribor University Library. His main research interests include processing of geometric data and compression of multimedia information. He authored more than 145 papers in scientific journals, the majority of them with the impact factor, 11 patents, and supervised 22 Ph.D. students. His hobby is radio-amateurism, where he operates under call-sign S58X. 5 GeMMA Activity Report 2016–2022 dr. Marko Bizjak Marko Bizjak finished the primary school Braslovče in 2006. During those years he successfully participated in various national-level competitions which allowed him to be awarded with the Zois scholarship. Four years later he finished the Gymnasium Lava in Celje and enrolled in Computer Science and Information Technologies study programme at UM FERI. He completed the programme in 2013 and continued his studies at the same programme on the master’s degree level, which he completed two years later. During his studies, he actively collaborated in GeMMA doing research work, for which he received the Andrej Perlach’s award, won a student paper competition (SPC) at ERK 2015 and was selected among five finalists of the IEEE Region 8 (Europe, Asia, Africa) SPC 2016. He started his first employment in December 2014 as a technical assistant at UM FERI. Next year he began his Ph.D. study, which he completed in 2019. He now works as a teaching assistant and researcher. His main research interests are remote sensing, environmental simulations and computational geometry, while his hobbies include volleyball, table tennis and football. Jan Breznar Jan Breznar was born on 6th February 1999 in Maribor. He finished the primary school of Sveti Jurij ob Ščavnici in 2014. From 2014 to 2018 he attended the Electrical and Computer School Ptuj, where he graduated as a computer technician. In the same year he enrolled in Computer Science and Information Technologies study programme at UM FERI. He finished his bachelor’s studies in 2021 and earned his bachelor’s degree in computer science. Currently he is pursuing a master’s degree in the same field. From 2019 he is also employed in GeMMA, where he works as a technical associate. His main research interests are GIS, remote sensing and web applications, while his hobbies are cycling, fishing, birding, hiking and reading. 6 GeMMA Activity Report 2016–2022 Matej Brumen Matej Brumen finished primary school in a town called Benedikt. After enrolling into Secondary School of Electrical Engineering and Computer Science he moved to his old hometown called Jurovski Dol. While he was studying there, he found out the passion for coding. After finishing the secondary school, he decided to pursue computer science career, so he enrolled into Faculty of Electrical Engineering and Computer Science, University of Maribor. Not knowing what he'd expect there he quickly found the study interesting and, therefore, finishing it became his main objective. During his 3rd year he was recruited by GeMMA, the laboratory on the same faculty and started his career there. Soon he graduated and obtained the title »dipl. inž. rač. in inf. tehnol.«. After graduating he enrolled MSc of computer science programme on the same faculty, from which he graduated in 2017. His main interest is exploring modern software engineering practices for minimal development overhead of the project teams, while also contributing code and ideas on the projects. His hobbies consist of playing the guitar and coding. Jernej Cukjati Jernej Cukjati finished the primary school in Prebold in 2010. After that he entered the Gimnazija Celje – Center and graduated in 2014. The same year he enrolled into the Computer Science and Information Technologies academic-degree study programme at the Faculty of Electrical Engineering and Computer Science (UM FERI), which he completed it in 2017. He continued his study on the master’s degree level and completed it 2 years later. During this study he also worked as a demonstrator. In 2019, he was employed as a young researcher in GeMMA and began his Ph.D. in computer science, with the focus of his research including the remote sensing data and data fusion. He also works as a teaching assistant. 7 GeMMA Activity Report 2016–2022 Tamara Golob Tamara Golob was born in Maribor in 1988. After finishing her primary school in Starše, she attended III. gimnazija Maribor. She was accepted at the Faculty of Law Maribor in 2007, where she got her bachelor’s degree. She started her career working in non-governmental sectors in various positions, where she has gained industry knowledge and interpersonal skills. She is a member of GeMMA since 2019, where she is working on several EU (H2020, Horizon Europe) and National funded projects, which include project management, coordination, networking, project and financial planning, administration, consulting, and reporting activities. Her hobbies are hiking, playing squash, and traveling. Gregor Horvat Gregor Horvat was born on 13th December 1997 in Maribor. He finished primary school Sladki Vrh in 2012. From 2012 to 2016 he attended Srednja elektro- računalniška šola Maribor, where he graduated as a computer technician. In 2016, he enrolled in Computer Science and Information Technologies programme at UM FERI. He received his bachelor’s degree in 2019 and applied for a master’s degree the same year. As of November 2022, he is employed at GeMMA as a technical associate and is working on his master thesis. His main research interests are machine learning, mobile application development and data compression algorithms. His hobbies include football, gaming and weightlifting. 8 GeMMA Activity Report 2016–2022 Štefan Horvat Štefan Horvat was born on 31st of December 1998 in Murska Sobota. He finished primary school of Tišina in 2013. From 2013 he attended secondary school SPTŠ Murska Sobota. After finishing secondary school in 2017, he enrolled in Computer Science and Information Technologies study programme at FERI UM. He received bachelor’s degree in 2020 and is currently pursuing master’s degree in the same field. Since October 2022 he is employed in GeMMA, where he works as a technical associate. His main research interests are machine learning, GIS, and forecasting. His main hobbies are running, cycling, astronomy and geography. Aljaž Jeromel Aljaž Jeromel finished the primary school Tabor I Maribor in 2010. During that time he regularly and successfully competed in nation-level competitions in math, logic, chemistry and physics. He attended the II. gimnazija Maribor between 2010 and 2014, where he started learning about computers and programming. Because of that, he enrolled into the computer science programme at UM FERI, which he finished in 2017. After graduation, he soon started the employment at GeMMA, firstly as a technical assistant, and, after obtaining the masters degree from the same faculty in 2019, as a teaching assistant. After a one-year pause from studies, he again enrolled into the Ph.D. programme Computer Science and Informatics at UM FERI in 2020. His main research interests include image compression, chain codes, 3D graphics and visualisation. His hobbies are football and playing computer games. 9 GeMMA Activity Report 2016–2022 dr. David Jesenko David Jesenko was born on 17th October 1990 in Celje. He attended the Primary School Šmarje pri Jelšah between 1997 and 2005. Four years later, he finished the Secondary School of Chemistry, Electrical Engineering and Computer Engineering in Celje. In autumn 2009, he entered the Faculty of Electrical Engineering and Computer Science (UM FERI) at the University of Maribor. He obtained a Bachelor’s degree and Master’s degree in 2012 and 2014, respectively. He started his first employment in September 2014 as a young researcher in GeMMA. He defended his Ph.D. thesis successfully in April 2018. During his study, he was also a guest researcher at Fraunhofer Ernst-Mach-Institut in Freiburg, Germany in February 2018. He has participated in several research, industrial and bilateral projects. His main research interests are Complex Networks, Evolutionary Algorithms and Machine Learning, while his main hobby is football. Domen Kavran Domen Kavran has obtained a bachelor's degree and master's degree in Computer science at UM FERI in 2018 and 2020, respectively. In 2020, he enrolled in the Ph.D. programme. Since 2018 he has been employed in GeMMA, where he started working as a technical assistant. His work involved web application development for various industrial projects. From 2020 onward, he has worked as a teaching assistant for web development and advanced algorithms undergraduate courses. His main research interests are time series analysis, change detection, pattern recognition, and advanced classification methods with artificial intelligence. 10 GeMMA Activity Report 2016–2022 dr. Štefan Kohek Štefan Kohek finished the primary school in Ljutomer in 2003. In 2007, he graduated at SERŠ in Maribor as an electrical-computer technician. He entered UM FERI in the same year and in the years 2010 and 2012 obtained B.Sc. and M.Sc. in computer science, respectively. During the study he was occupied by a sole proprietorship as a programmer. In 2012, he became a technical assistant in the CGAI lab at UM FERI and from 2013 onward he is employed as a teaching assistant. By joining the former laboratory to GeMMA in the year 2016, he has also joined the new laboratory. In 2019, he obtained a Ph.D. degree in computer science. His current research interests include computer graphics, tree growth simulation, optimization techniques, parallel computing, and remote sensing data. Denis Kolednik Denis Kolednik finished the primary school of Benedikt in 2004. Later he attended the Secondary School of Electrical Engineering and Computer Science Maribor – Technical gymnasium and finished in 2008. He started his study path at the UM FERI in 2008. In 2011, he finished his Bachelor studies (1. bologna degree) of Computer science and information technologies. In 2013, he finished his Masters studies (2. bologna degree) of the same study programme and got his first employment as a technical assistant in GeMMA in August 2013. Half a year later he started his research career as a researcher at GeMMA. From 2014 to 2017 he was also a teaching assistant. His current research topic is Geographic Infromation Systems. His main hobby is bicycling. During this study he also worked as a demonstrator. In 2019 he was employed as a young researcher in GeMMA and began his Ph.D. in computer science, with the focus of his research including the remote sensing data and data fusion. He also works as a teaching assistant. 11 GeMMA Activity Report 2016–2022 dr. Simon Kolmanič Simon Kolmanič finished the primary school Ormož in 1987. He continued his education at the Secondary School of Natural Sciences and Mathematics Ptuj, which he completed in 1991. During that time, he got his first computer Commodore 64, and discovered the beautiful world of programming and computer graphics. Consequently, he continued with his education on UM FERI where he finished the study in 1996 with a B.Sc. degree. In the same year, he joined the CGAI lab in the same institution as a technical assistant. In 1997 he started his M.Sc. study, which he finished in 1999. In the same year, he also started working as a teaching assistant. In 2000, he started his Ph.D. study, which he finished in 2005. In 2016 the CGAI lab joined GeMMA, where he continues with his work. During his work in CGAI lab, he started working in the field of computer animation, which is still one of his main interests. He is also interested in the virtual and augmented realities and computer simulations of vegetation growth and ecosystems. Between 2006 and 2011, he was a member of the Electronic Communication Council of the Republic of Slovenia. Currently, he is an assistant professor and is giving lectures in the field of algorithms, computer graphics, and computer animation. From the mid of eighties, when he got his first camera, he has been in love with photography, which is his main hobby to this day. He is also active as a volunteer firefighter in Voluntary Fire Department Ormož, where he was also a fire chief between 2003 and 2011. From time to time, he can also be found as an actor on a stage with some quite successful performances behind him. 12 GeMMA Activity Report 2016–2022 dr. Bogdan Lipuš Bogdan Lipuš completed the primary school in Oplotnica in 1990. In 1994, he successfully finished SERŠ in Maribor. After that, he entered UM FERI and graduated in computer science in May 2000. In his student days in December 1999, he started to work as a technical assistant in the CGAI lab. In 2000, he got a position as a young researcher and entered the M.Sc. programme. After obtaining M.Sc. in computer science in 2003, he continued his study and received his Ph.D. in computer science in 2005. For two and a half years he worked as a software developer in Hermes Softlab. Returning from industry, he started to work in GeMMA. He participated in several applicative and scientific research projects. Since 2015, he was elected as an assistant professor of computer science. His research interests include point cloud processing, remote sensing, computer graphics, processing of geometric data, data compression, and image processing. Currently, he works mostly on industrial projects. His hobbies include cooking, gardening, walking, running, and cycling. Luka Lukač Luka Lukač was born on 13th of February 2000 in the small town of Murska Sobota. After finishing primary school of Bakovci in 2015, he crossed the Mura River in order to enroll at Franc Miklošič High School in Ljutomer, which became his second home until 2019. During that time, he grew fond of computers, mathematics and physics. Therefore, in 2019, he decided to start his studies at FERI UM in Computer Science and Information Technologies study program. Besides obtaining his bachelor’s degree in 2022, he started working in GeMMA as a technical associate, where his research interests are focused on GIS and data compression. His main hobbies are participating at trivia quiz events, doing several sports, watching splendid movies and reading great books. 13 GeMMA Activity Report 2016–2022 dr. Niko Lukač Niko Lukač is an Associate Professor in the field of Computer Science at Faculty of Electrical Engineering and Computer Science, University of Maribor, Slovenia. He completed his Ph.D. study in Computer Science at University of Maribor in 2016 under Young Researcher Ph.D. study scholarship. During his Ph.D. studies he was also a visiting Ph.D. student at German Aerospace Center (DLR) and Heidelberg University, in 2014 and 2016, respectively. Soon after completing his Ph.D. study he was habilitated as Assistant Professor in 2017, and as an Associate Professor in 2022. His research focus is in the following areas: geospatial data analytics, simulations and modelling, parallel computing, and applied artificial intelligence. In the past years he has co-authored several journal papers indexed in Science Citation Index (SCI), international conference papers, book chapters, and received a US patent grant. In the given timeline he has also successfully coordinated various R&D projects at national and international level. As an Associate Professor at UM FERI, he is also active pedagogically, by providing thesis supervision and lectures for undergraduate and postgraduate students. During the timeline 2019-2022 he was an executive committee member of the European Umbrella Organisation for Geographic Information (EUROGI). From 2018 to 2022 he also served as the Section Editor for the Computer Science field at the open access journal Data in Brief, while during the years 2020-2021 he also served as a Topic Editor for ISPRS International Journal of Geo-Information. In 2022 he also took part in expert evaluation process of Innovation Action (IA) type of proposals for the European Commission, under the Horizon Europe programme. In 2019 he has received recognition at the University of Maribor for outstanding research achievements. 14 GeMMA Activity Report 2016–2022 dr. Domen Mongus Domen Mongus was born in Slovenj Gradec in 1982. He spent his youth in a small village beneath the Carinthian Mountains, by the name of Podgorje. After the elementary and secondary schools, he left for Maribor to study computer science at UM FERI. In 2007, he started to work in GeMMA. He defended his diploma thesis in 2008 and started working as a Young Researcher, funded by the Slovenian Research Agency. He concentrated his work on Environmental Intelligence, joining the fields of data fusion, remote sensing data processing, geometric pattern recognition, and artificial intelligence. In 2012, he completed his Ph.D. and became an assistant professor at UM FERI. In the same year, he received an award for research excellence at UM FERI. In 2013, he received an award for pedagogic excellence at UM FERI, while the Slovenian Research Agency awarded him for Exceptional Achievement in Science in 2014. In 2015, Slovenian National Radio and Television, Val 202, named him as “The Name of the Week”. In the same year, he received the highest award in the field of Information Society in Slovenia for ongoing work. He was also named as Young Scientist of Danube region by Danube Region and Central Europe and Austrian Federal Ministry for Science, Research and Economy. In 2018, he received the highest institutional academic award for exceptional contributions to scientific and pedagogical reputation and excellence of University of Maribor, and was awarded for exceptional research achievements in 2019 at University of Maribor, Faculty of Electrical Engineering and Computer Science. From 2008 to 2012, Domen Mongus was a member of the Executive Committee of ACM Slovenia. From 2013 to 2019, he was a member of the Executive Committee of European Umbrella Organization for Geographic Information (EUROGI) and is a member of Executive Committee of GISIG from 2020. 15 GeMMA Activity Report 2016–2022 Andrej Nerat Andrej Nerat was born on 6th October 1980 in Maribor. Since early childhood he developed interest in computers, first an old ZX Spectrum, later PCs. After he finished primary school in Ceršak and later Šentilj v Slovenskih goricah in 1995, he continued his education at SERŠ in Maribor. In 1999, he first entered UM FERI, where he has been since then. He spent the first five years there as an undergraduate student in computer science. After graduation in 2004 he joined the CGAI lab. He has worked there as a technical assistant. In 2016, he joined GeMMA to continue working at the same position. Sašo Pečnik Sašo Pečnik was born in Maribor in 1985. He finished primary school in Miklavž in 2000. Four years later he finished SERŠ in Maribor as a Computer Technician and then entered UM FERI, where he graduated in Computer Science as the best of his class in 2007. Two years later he received his MsC. In 2008, he was noticed by his professor and mentor Borut Žalik who gave him his first employment in GeMMA. He entered a Ph.D. study program of Computer Science and was promoted to a researcher in 2010. In 2014, he became a teaching assistant at UM FERI. Between 2014 - 2015, he worked as part-time researcher at the Company Lineal d.o.o. in Maribor. Along with the Lineal team, he was awarded at CITA Smart Collaboration Challenge 2014 in Dublin with the 1st place for the project RO3D-SMART. His main research interests are processing and visualization of LiDAR data along with computer geometry, computer graphics, CAD and cloud computing, while his hobbies are running, football and travelling. 16 GeMMA Activity Report 2016–2022 dr. David Podgorelec David Podgorelec finished primary school Maks Durjava in Maribor and Secondary School of Natural Sciences and Mathematics (the present-day II. gimnazija Maribor) in 1982 and 1986, respectively. He graduated in computer science at the Technical Faculty (predecessor of UM FERI) in 1993 and found his first employment as a programmer in MIPS d.o.o. in Maribor in 1994. In autumn 1995, he got a position of a young researcher at UM FERI, but soon changed it to a teaching assistant position. After four years in the CGAI lab, he moved to GeMMA in 2000 as one of its three original members. He obtained M.Sc. and Ph.D. in computer science from UM in 2000 and 2002, respectively. During 2004–2005, he spent 7 months at the University of Luton (current University of Bedfordshire) in the United Kingdom as a research fellow. In subsequent 10 years, he worked as assistant professor of computer science at UM FERI. Between April 2012 and February 2015, he was a head of the Media Communication Institute at UM FERI. In October 2015, he completed his assistant professor career, but he returned to GeMMA in March 2016 as a researcher. He has remained in this position until now, with the exception of the last third of 2018 when he was employed at the University of Ljubljana. In his prime, he used to climb mountains and play football as hobbies, which he traded for cycling, mushroom picking and dog-walking in his mature years. 17 GeMMA Activity Report 2016–2022 Blaž Repnik Blaž Repnik finished the primary school Črešnjevec in 1998. Four years later he finished SERŠ in Maribor. In 2002, he entered UM FERI where he got the bachelor degree in 2007. He started working in GeMMA in 2006 as a technical assistant. He entered the Ph.D. study programme of Computer Science and was promoted to researcher position in 2013. His main research interests are GIS, dynamic systems and 3D graphics. dr. Damjan Strnad Damjan Strnad graduated from computer science and informatics at UM FERI in 1998. He upgraded his education through the M.Sc. in 2000 and Ph.D. in computer science and informatics in 2006. For his study excellence he received the university chancellor‘s award. In 1997, he was employed as a technical assistant in the CGAI lab. He continued working in CGAI as an assistant during 1998-2007. Since becoming the assistant professor in 2007 and the associate professor in 2012, he is giving lectures in the field of computer science, particularly computer graphics and artificial intelligence. The latter is also his main research interest. He has been working as a supervisor for several diploma candidates and is currently mentoring a Ph.D. candidate. He joined GeMMA at the start of 2016. 18 GeMMA Activity Report 2016–2022 Niko Uremović Niko Uremović finished his primary schooling in Maribor in 2013. Four years later, he finished II. gimnazija Maribor. He obtained bachelor’s degree in 2020 on University of Maribor, Faculty of Electrical Engineering and Computer Science. He started his first employment in September 2020 as a young researcher at UM FERI. During his studies at secondary school he was a recipient of the Zois scholarship for outstanding students. His main research interests are machine learning and IoT, while his main hobbies encompass reading, spending time with family and training his dogs. Dino Vlahek Dino Vlahek was born in Čakovec in 1992. He spent his youth in a small village in Međimurje County, by the name of Čukovec. After finishing the primary school called OŠ Sveta Marija in 2006 and, four years later, the Gymnasium in Čakovec, he went to obtain the bachelor's degree in computer science at The Polytechnic of Međimurje in Čakovec. After earning it, he worked in the private sector as a computer programmer for a few years. In 2016 he entered the M.Sc. study program of informatics and technologies of communication at UM FERI. He defended his master's thesis in 2018 and was recruited by GeMMA, building his career as a researcher. At this point, he also started his Ph.D. study in Computer science and informatics at UM FERI. His current research interests are feature learning, data analytics, and model interpretation. 19 GeMMA Activity Report 2016–2022 Mitja Žalik Mitja Žalik finished the primary school Kamnica in 2013. In the next four years, he attended grammar school II. gimnazija Maribor, where he started to learn programming. During that time, he successfully competed in several national competitions, mainly in physics, mathematics, logic and programming. In 2017, he qualified and participated in Central European, Balkan and International Olympiads in Informatics. In the same year, he enrolled in the Computer Science and Information Technologies programme at the UM FERI. In 2019 he started to work in GeMMA as a technical assistant. He obtained a bachelor’s degree in 2020. Two years later, he finished his master’s degree and became a teaching assistant. While studying, he was a member of the team that successfully participated in various programming competitions (ICPC Slovenian Programming Contest - UPM, ICPC Central Europe Regional Contest – CERC and IEEEXtreme). Currently, he is enrolled into a doctoral degree study program. His research interests include edge computing, geospatial data processing and convolutional neural networks. Aljaž Žel Aljaž Žel was born on 22nd of June 1999 in Maribor. He finished primary school Prežihovega Voranca Maribor in 2014. From 2014 to 2018 he attended II. gimnazija Maribor. After finishing secondary school he started his studies at UM FERI. In 2021, he received his bachelor’s degree. Currently he is enrolled in computer science master's degree. In secondary school and university he received Zois scholarship for outstanding achievements. In 2021 he started to work in GeMMA, where he currently works as a technical associate. His main research interests are GIS, machine learning and computer graphics. His main hobby is running. 20 GeMMA Activity Report 2016–2022 Denis Žganec Denis Žganec finished his primary school in Štrigova. From 1999 to 2003 he attended the secondary school of Technical School in Čakovec, where he graduated as a computer technician. In 2005, he was accepted to UM FERI in order to study computer science. He obtained bachelor’s degree in 2012, and he is currently finishing his master’s. From 2008 he is also employed in GeMMA, where he works as a technical associate. His work is mainly concentrated on the development of GIS. Part-time members YEARS ACTIVE FROM dr. Simon Jurič 2014 dr. Krista Rizman Žalik 2004 21 GeMMA Activity Report 2016–2022 Past members YEARS ACTIVE FROM TO Robi Cvirn 2015 2018 Roman Čuk 2013 2013 dr. Vid Domiter 2004 2011 dr. Simon Gangl 2010 2014 dr. Matej Gomboši 2000 2006 dr. Denis Fekonja 2012 2017 Valentin Kerman 2018 2021 dr. Gregor Klajnšek 2001 2010 Primož Kovačič 2012 2013 dr. Sebastian Krivograd 2000 2009 Žiga Leber 2015 2018 Simon Lušenc 2013 2014 Renato Mikša 2009 2010 Denis Obrul 2006 2015 Amadej Pevec 2015 2018 Boštjan Pivec 2004 2015 Damjan Roškar 2014 2015 dr. Bojan Rupnik 2006 2013 dr. Gregor Smogavec 2008 2014 Tadej Stošić 2018 2022 dr. Denis Špelič 2004 2018 Jan Tovornik 2018 2019 dr. Mirko Zadravec 2001 2008 dr. Eva Zupančič 2017 2020 dr. Danijel Žlaus 2011 2022 22 GeMMA Activity Report 2016–2022 ST Slovene National Projects International Projects Bilateral International Projects Industrial Projects PROJEC 23 GeMMA Activity Report 2016–2022 Slovene National Projects 24 GeMMA Activity Report 2016–2022 Computer Systems, Methodologies and Intelligent Services - Programme Funded Unit Programme Funded Unit (PFU)– Computer Systems, Methodologies and Intelligent Services started in 1999 and is implemented in five-year funding periods (with the exception of the last period, which lasts 6 years). Research Financed by goals in the PFU are adapted to current trends in computer science. Since ARRS – Slovenian Research Agency the begining, researchers from most laboratories of the UM FERI Institute (contract P2–0041) of Computer Science participate in PFU. In the last two funding periods, which are briefly described in this chapter, the participating laboratories Duration are: 1999 to 2025 • Computer Architecture and Languages Laboratory; • Laboratory for Geospatial Modelling, Multimedia and Artificial Intelligence (GeMMA); Aditional information • Laboratory for Heterogeneous Computer Systems; http://p2-0041.feri.um.si/2015_2019/ • Programming Methodologies Laboratory; https://p2-0041.feri.um.si/ • System Software Laboratory. Period 2015-2019 The research examined the common features and laws of unstructured and heterogeneous massive data sources and flows we encounter on a daily basis in computing and informatics (e.g. World Wide Web, Earth surface data acquisition systems, biomedical systems). Their scope, dynamics and diversity offer many research challenges, the primary goal of which is to unify their processing at the appropriate level of abstraction. For this purpose, we have broken down individual data sources and streams, which are usually immersed in strong information and instrumental noise, into basic semantic building blocks (symbols), which enabled their efficient noise reduction, structuring and alignment. Advanced data enrichment methods were implemented in the form of weakly merged services, the orchestration of which led us to a wide range of user applications. To this end, we have integrated key research paradigms into a three-tier service architecture, where the first level took care of domain-specific data resource management and interoperable access to second-level services. This level was focused on data enrichment, and we paid a lot of attention to obtaining basic data building blocks. We focused primarily on the development of two recently proposed paradigms: algebraic formalization of attribute filters based on mathematical morphology, and analysis of hidden components. The first paradigm allows accurate estimation of sample properties by selectively and fully automatically adapting the required geometric structures to input data sets, while the second paradigm exploits the time-space dependencies of data building blocks (symbols) to separate composite data streams from different sources. 25 GeMMA Activity Report 2016–2022 We looked for heuristic knowledge about the characteristics of the obtained basic data building blocks with the help of machine learning algorithms and connected them into multi-meaning sets through their mutual relations. The last level of our architecture was the application level, where we used the services of the second level and show their universality and interoperability in very different areas. Typical examples include the detection of irregular muscle contraction with the help of non-invasive surface electromyograms and the assessment of changes in the earth's surface due to landslides, water or wind erosion. Both of these applications address current socio-economic challenges and are linked by the rapidly rising costs of demographic and climate change. Extensive information support based on the collection and credible interpretation of verifiable data is crucial for the effective adoption of strategies at national and European level. The proposed development of computer algorithms has enabled more efficient, reliable and faster processing of existing databases in these areas, and thus greatly supported many other scientific fields. Period 2020-2025 Growth in Internet of Things (IoT) investments, mass data analysis, and artificial intelligence has spurred the development of digital copies of real-world entities in the form of digital twins. Such cyber-physical systems offer advanced monitoring, data analytics, and prognostic capabilities, making them a new trend in computing. Gartner ranks them among the top 10 technologies in 2019, with an expected 37% annual growth rate from the current $ 2 billion to $ 15 billion in 2023 and $ 26 billion in 2025. With the ability to anticipate potential problems and find optimal solutions, such digital twins can offer significant help in treating patients and help reduce risks and increase treatment effectiveness. Nevertheless, today the use of digital twins is limited mainly to highly controlled environments and smart machines. However, the development of technologies to mimic more complex systems related to the functioning of the human body still faces the following important challenges: • Processing the set of heterogeneous data streams needed to learn the behavior of the observed system requires significant improvements in methodologies for automatic data alignment and structuring; • Existing methods of merging medical data and learning characteristics are still focused mainly on the isolated processing of individual data sources. This requires the development of new methods that will be able to make better use of their complementarities; • Linking biomedical measurements with environmental and lifestyle factors, which is essential for the transfer of laboratory observations to real environments, requires significant advances in methods for extracting contextual characteristics; • Methodologies for monitoring living microhabitats need to be improved, as high dispersion of environmental sensors creates large spatial and temporal gaps in the obtained informations; • The need to personalize digital twins requires the optimization of dynamic models and their adaptation to the observed persons, which exceeds the capabilities of modern optimization algorithms. 26 GeMMA Activity Report 2016–2022 As part of the proposed work program, we intend to upgrade our previous research work and address the described challenges with the aim of implementing a digital twin that will be able to mimic the functional parameters of the human nervous and muscular system in the real environment. Due to the general aging of society, neuromuscular diseases are becoming an important health risk and a leading cause of incapacity for work. The costs associated with such diseases in Slovenia exceed 2% of GDP. The program group brings together leading experts in the processing of neuromuscular signals and semantic data, development of methods of temporal and spatial analysis and implementation of optimization algorithms that will implement the proposed program in a co-creative way based on a focused iterative work plan. Picture below shows the developed web platform for the digital twin, where a given test subject who is anonymized has several attached sensors (e.g. heartbeat), which can then be visualized and analysed by the platform. The platform enables fusion of time series sensor data from the test subject, with time series sensing data from stationary sensors (e.g. weather stations) as well as Earth observation (EO) data (e.g. satellite imagery). 27 GeMMA Activity Report 2016–2022 Morphological Operators for Pattern Recognition in Large Point Clouds The advanced technologies of laser scanning with their accuracy, speed and resolution, have revolutionized the field of Earth observation. Financed by The amount of information contained within 3D point clouds has introduced the recognition of geometric structure as the most important ARRS – Slovenian Research Agency (contract computational challenge of this decade. Developing new solutions J2–5479) requires coping with irregular point distribution, the lack of topology Duration and their sheer size that often exceeds the capabilities of modern computer systems. Using the known concepts that were developed for 2013 to 2016 pattern recognition in raster data leads to inefficient algorithms that require intensive user interaction and additional information about the geographical areas. Partners The proposed project’s intention is to research a new methodology for UM Faculty of Electrical Engineering and recognizing 3D geometrical structures, monitoring their dynamics, and Computer Science, Geodetic Institute of detecting events within large point clouds, as acquired from scanning Slovenia the Earth’s surface by applying contemporary findings of mathematical morphology. Although, mathematical morphology is considered to be a young mathematical theory, its quantitative arithmetic of shape description offers great expressional strength. Morphological operators are derived from the set theory and extended by using the concepts of geometry, topology, probability, and statistics, and are completely adapted for digital and parallel processing. The recently developed algebraic formalization of scanning morphology offer a spatially-dependent, selective, and completely automatic adaptation to the geometrical structures of input data. These theoretical foundations offer the possibility of developing an efficient pattern recognition methodology, where adaptation to the temporal domain would allow a quantitative presentation of events and a description of the dynamics. The efficiency of the developed method would be demonstrated with by two uses: • Recognition of geomorphological process kinematics and • Monitoring tree development in Slovenia. For the purpose of recognizing the kinematics of geomorphological changes (such as landslides) it is intended to develop an automatic method for ground recognition within 3D point clouds, and the construction of a digital elevation model that would be more accurate and time efficient without the need for users to set parameters. Such a procedure would allow for the detection of changes in the terrain and evaluate the volume, mass and speed of moving earth masses over large geographical areas (whole of Slovenia) with high resolution (under 0.5m) and accuracy (over 90%). Similar accuracy can be expected regarding (ii) monitoring tree development, where a new method for recognizing single trees would be developed. This method would estimate the number of trees within a respected area and provide the geographical positions, heights and volumes of tree-crowns. It would measure growth of a single tree, wood biomass growth by cyclical data acquisition, and develop a predictive simulation of their development. 28 GeMMA Activity Report 2016–2022 The precision of the proposed uses would be tested by on terrain measurements, while the construction of a digital elevation model of Slovenia will demonstrate their computational efficiencies. In this way, the national project for surface scanning of Slovenia with LiDAR technology would be supported directly. The results of our research will be published in the most distinguished international journals, and regularly presented to the Slovenian public by organizing symposiums and workshops. It will also promote our products abroad by attending international conferences. Figure 4: Decomposition of a grid achieved by (a) progressive filtering of g at increasing scales s, where a response vector ∆(g)[p] is assigned to each grid-point, estimated from (b) the input LiDAR point-cloud. (Source: own) 29 GeMMA Activity Report 2016–2022 Assessment and Optimization of Planning and Implementation of Tending Young Forest in Slovenia The decrease in the realization of planned silvicultural treatment (tending) and concurrent increase of regeneration fellings may lead to a Financed by long-term decrease of quality and stability of the private and state forest and, at the same time, its capability to provide the ecosystem services. ARRS – Slovenian Research Agency, MKGP In the period 1993-2011, the realization of planned tending measures RS (contract V4–1420) was around 58 %. The decrease of tending activities was especially pronounced in private forests, where only one-third of planned tending Duration was implemented. This decrease is, on the one hand, a result of socioeconomic changes and, on the other hand, a result of the decline in the 2014 to 2017 state subsidies for tending. The decrease in the realization of tending measures could also be attributed to the prevalence of continuous cover silvicultural systems used in Slovenia, where education of young forest Partners is done mostly by the appropriate canopy cover. UM Faculty of Electrical Engineering and Computer Science, UL Biotechnical faculty, One of the demands connected to the Slovenian state and European The Slovenian Forestry Institute Union subsidies for tending is a need to separate between tending for increasing the profitability of the forest and the tending that strengthens and preserves long-term ecosystem services of the forest. This project aimed to review existing tending standards and to develop a tending strategy, especially for damaged forests, since in February 2014, almost 400.000 ha of Slovenia forests were damaged by an ice storm. Practical cases have shown that the damage caused by irresponsible salvage activities could be greater than the harm caused by the storm. During the salvage, it is important that we give special attention to fine-tuning of salvage logging and biological restoration, processes of secondary succession, setting priorities, and taking into account the recommendations of good practice in silviculture. Our role in the project was to provide the needed support by the decisions about using natural regeneration versus planting in highly damaged younger stands. For that purpose, the secondary succession model ForestMAS has been used, which is based on Ellenberg ecological values. With the help of ForestMAS and its ability to interfere with the forest composition by generating clear-cuts and removing or planting individual trees, the regeneration of damaged areas could be studied and thus help to evaluate the tending models. 30 GeMMA Activity Report 2016–2022 Figure 5: Generating the clearcut regions (gaps) into dense forest countryside and the ground view of the regenerating area. (Source: own) Figure 6: Forest gap regeneration through 10, 30, and 50 years, respectively, depending on given seed trees. (Source: own) 31 GeMMA Activity Report 2016–2022 Algorithms of Ecosystems Dynamics Modelling with Methods of Mathematical Morphology and Lattice Theory Sustainable management of the environment is a major challenge facing not only Slovenia, but also the entire mankind. Large ecosystems, Financed by especially forests, play a major role when addressing this task, having critical impact on the quality of life and obvious social-economic benefits ARRS – Slovenian Research Agency (contract for the society. Systematic and complete monitoring of the evolution of J2–6764) such ecosystems is extremely difficult, up to now even impossible, due Duration to their vast geographic scales and huge amounts of their miniature basic elements. Only recent advances in remote sensing technologies 2014 to 2017 that have revolutionized the area of Earth observations provide us with possible insights into the dynamics of such ecosystems. Sophisticated satellite observation systems from the Copernicus program and state-Partners of-the-art laser scanning technologies like LiDAR, allow for periodical monitoring of large geographical areas with high enough resolution and UM Faculty of Electrical Engineering and precision to distinguish the smallest basic elements of ecosystems, Computer Science, The Slovenian Forestry such as trees, undergrowth, and shrubs. However, the huge amounts Institute of heterogeneous and complex data they acquire remains a major challenge for the future as contemporary software solutions are incapable to deliver data analytics in a systematic, organized manner. Before a holistic information space for efficient management of large ecosystems can be developed, major issues have to be addressed, regarding integration of heterogeneous Earth observations data, implementation of relevant analytic tools for their processing and, finally, relevant models of their dynamics. The proposed project meets these challenges by introducing a new paradigm for data integration based on the decomposition of heterogeneous Earth observations into the contained basic semantic elements, their fusing and enrichment with complementary information from within different data types, and their inter-linking into a complex network. Through advanced concepts of mathematical morphology, formalizing arithmetic of shapes for sophisticated pattern analysis, the decomposition of specific data types and the recognition of the basic ecosystems’ elements will be achieved. Their geometric features will be used to determine their social status, and consequently the likelihood of their mutual influence. These will be represented with a complex network, enabling us to develop a wide range of new algorithms based on up to now unexploited mathematical and analytical methods at such large scale. This new type of data analytics will be derived primarily from methodological studies of partially ordered sets based on lattice theory and statistical-topological features based on the theory of complex networks. 32 GeMMA Activity Report 2016–2022 Such fundamental shift in the design of the pattern recognition algorithms will provide the thoughtfully required capabilities for the development of new approaches to recognition of complex structures, composed of multiple basic elements, while comparison of complex networks will allow for systematic monitoring of their evolution. Hence, the foundation for recognizing interactions between the basic elements will be established, giving us the framework for modelling dynamics of large ecosystems. While in-situ measurements will be used to validate these algorithms, a study of forest dynamics due to the competition of trees for accessing resources and leaving space will provide the proof of concept. All the developed methods will be implemented in the form of weakly coupled service for this purpose and integrated into an existing platform for geographic data management and processing. A user-friendly environment for services orchestration and execution of analytic scenarios on-demand will be provided to experts in a form of end-user application. Fogure 7: Visualized LiDAR point cloud of a forest area decomposed to single trees. (Source: own) 33 GeMMA Activity Report 2016–2022 InfraCloud – An Innovative Process for Creating Digital Models of Built-in Infrastructure Using a Cloud of Points GeMMA acted as an external contractor in this project. The aim was the development of semi-automatic algorithms for the recognition of Financed by built infrastructure and creation of virtual models for further design, reconstruction of built infrastructure or managing and maintaining EU (ERDF); MIZŠ RS infrastructure. Duration Accurate recognition of built infrastructure and the production of digital models enables integrated integral planning and management, and 2016 to 2018 more efficient management and safer execution of interventions in the environment. Partners The application recognizes dominant points from input LiDAR data, defines built infrastructure element, and produces a digital model UM Faculty of Electrical Engineering enriched with geometry and object attributes (BIM model). The model and Computer Science, CGS plus is then exported into open-source format in order to be exchanged with d.o.o., UM Faculty of Civil Engineering, other software solutions. Transportation Engineering and Architecture Additional information http://cgsplus.si/projekt-infracloud/ Figure 8: Created digital enriched model of build infrastructure from LiDAR data with Infracloud. (Source: own) 34 GeMMA Activity Report 2016–2022 PAKT Architecture – IoT Based Machine Learning Model for Predicting Electricity Consumption GeMMA acted as an external contractor in this project. The focus was on the analysis and machine learning architectures for predicting electricity production and consumption from real-time internet of things (IoT) Financed by data and, accordingly, balancing the electricity flows. For this purpose, EU (ERDF); MGRT RS extensive study of the existing technology for predicting electrical energy networks’ workloads was conducted and key strongpoint and good practices were identified. Based on a technological solution for the Duration use of machine learning methods in predicting electricity consumption and / or load on the electricity, the grid was designed at a conceptual 2016 to 2018 level and tested in as a proof-of-concept. Partners UM Faculty of Electrical Engineering and Computer Science, Inea d.o.o., A1 Slovenija d.d., Borzen d.o.o., Elpros d.o.o., Geodetski zavod Celje d.o.o., Igea d.o.o., Iskraemeco d.d., Metronik d.o.o., Seltron d.o.o., Semantika d.o.o., Sipronika d.o.o. Figure 9: Real-time energy production predictions in dedicated GIS. (Source: own) 35 GeMMA Activity Report 2016–2022 GOSTOP – Building Blocks, Tools and Systems for the Factories of the Future The aim of the proposed GOSTOP program was to accelerate the development of the Factories of the Future concept in Slovenia and to Financed by provide solutions to the current needs of Slovene industry, where some companies have already started to introduce this concept into their EU (ERDF); MIZŠ RS production facilities. Duration In GOSTOP, 13 companies and 6 research organizations with compatible research and development programs in the Factories of the Future 2016 to 2020 area joined forces to advance the concept. Considering the Smart Specialization Strategy of Slovenia prepared by SVRK (Government Office for Development and European Cohesion Policy) and the priorities Partners of the Factories of the Future roadmap under Horizon 2020 prepared UM Faculty of Electrical Engineering by European Factories of the Future Research Association (EFFRA), we and Computer Science, UL Faculty of have identified 4 areas in which decisive breakthroughs can be achieved Electrical Engineering, UL Faculty of in Slovenia in the near future: Mechanical Engineering, UL Faculty of • Control technologies; Computer and Information Science , • Tooling; Institute Jožef Štefan, TECOS, Kolektor • Robotics; d.o.o., Inea d.o.o., Metronik, Hidria • Photonics. Rotomatika d.o.o., Yaskawa Slovenija d.o.o., Podkrižnik d.o.o., Nela d.o.o., This means that in GOSTOP we combined most of the horizontal Cosylab d.d., L-TEK d.o.o., Špica fields pinpointed by the Smart Specialization Strategy of Slovenia international d.o.o., Optotek d.o.o., LPKF documents for the Factories of the Future area. In all of these fields d.o.o., Fotona d.o.o. we determined the most promising research topics that are interesting for Slovene industry, where the necessary knowledge in Slovene research organizations exists and identified synergies between them. Additional information We combined several value chains in the program within which new https://www.gostop.si/ products can be developed. This way, the competitiveness of Slovenian industry was improved significantly. On the one hand, GOSTOP includes the development of new products and breakthrough technologies by agile SMEs (small and medium-sized enterprise). On the other hand, we advanced the overall Factories of the Future concept, which lead to integrated systems that can be used by large Slovenian companies to optimize their production and develop new products with high added value. An example of such a product included in GOSTOP is the vision of a turnkey factory. The success of GOSTOP has contributed to raising both the added value and the export volume of the participating companies and Slovenian industry at large. 36 GeMMA Activity Report 2016–2022 EkoSmart – Eco System of the Smart City EkoSmart addressed the “Smart cities and communities” priority subarea within the Slovenian Smart Specialisation Strategy (S4) priority area of “Healthy working and living environment“. It was selected for Financed by funding within the ERDF and MIZŠ RS public call (2016) for proposals “to EU (ERDF); MIZŠ RS support Research and development projects (TRL 3-6)”. The consortium of 12 innovative companies and 12 top research Duration institutions was involved in EkoSmart implementation. The programme combined 6 interdisciplinary R&D projects (RDP1 to RDP6) in order to 2016 to 2019 develop a smart city ecosystem with all support mechanisms needed for efficient, optimized and gradual integration of individual areas into a unified and coherent system of value chains. 4 RDPs focused on three Partners key domains of smart cities: UM Faculty of Electrical Engineering and • Health (RDP4 and RDP5); Computer Science, Marand d.o.o., Jožef • Active living (RDP3); Stefan Institute, URI-Soča, Špica d.d., • Mobility (RDP2). UKC Ljubljana; UL Faculty of Electrical Engineering , UL Faculty of Computer RDP1 introduced self-configurable, self-integrating, self-optimizing, and Information Science, Inova IT d.o.o., flexible and adaptable universal smart city ecosystem architecture Elgoline d.o.o., Nela d.o.o., SRC d.o.o.; with capability of simple addition of modules, while RDP6 focused Cosilab d.d., ZD Adolfa Drolca, Iskra d.d, on development of prototype solutions and their testing in relevant RC-IKTS d.o.o., Telekom Slovenije d.d., environments. UL Faculty of Medicine, Robotina d.o.o., Alpineon d.o.o., UL Faculty of Sport, GeMMA participated in work package WP4 (“Digital support, data, and Klinika Golnik, Anton Trstenjak Institute acquisition of new knowledges”) of RDP4 (“E-health and mobile health”), of Gerontology and Intergenerational where it autonomously implemented two tasks – T2 (“Machine learning Relations, Medis d.o.o., National Application for Discovering States of Illness and for Intelligent Support Institute of Public Health to Their Medical Treatment” ) and T3 (“Enriching Medical Knowledge with Geographic Data” ). Within T2, different machine learning approaches were analysed, and Additional information six of them were then integrated into innovative data mining tool. The http://ekosmart.net/en/ekosmart-2/ selection consisted of neural networks, decision trees, clustering, support vector machines, linear regression, and random forest algorithm. The developed tool enables iterative execution in multiple loops. 37 GeMMA Activity Report 2016–2022 Typically, (hierarchical) clustering is performed in the early iterations, while the chosen attributes are predicted separately in each cluster later on. The tool also provides the visualisation of multidimensional data module for purposes of visual analytics. Within T3, the data analytics tool was upgraded with geospatial and other georeferenced data to provide functionalities of finding correlations between the medical and geographic attributes from heterogeneous data sources. The developed GIS extends the functionalities of visual analytics into the geospatial domain. After the completion of RDP4.WP4.T2 and RDP4.WP4.T3 goals, GeMMA also participated in RDP6 (Solution prototypes), where the developed data analytics and visualisation subsystem was made ready for integration into the common EkoSmart ecosystem and validation in a relevant environment. 38 GeMMA Activity Report 2016–2022 IQ DOM – Intelligent Home of a New Generation Based on Smart Devices and Wood IQ HOME (Sl IQ DOM) addressed the “Smart buildings and homes, including wood chain” priority subarea within the S4 priority area of “Healthy working and living environment“. It was selected for funding Financed by within the ERDF and MIZŠ RS public call (2016) for proposals “to support EU (ERDF); MIZŠ RS Research and development projects (TRL 3-6)”. The consortium of 26 partners from various technological areas was Duration involved in IQ HOME implementation. The programme combined 25 interdisciplinary R&D projects (RDP1 to RDP25) in order to provide 2016 to 2019 advanced technological solutions implementing a new paradigm of an integral, green, mostly wood-based intelligent home adapted to resident’s needs with extensive use of non-invasive artificial intelligence. Partners Eight RDPs considered the building itself (the value chain of “Advanced UM Faculty of Electrical Engineering buildings with wood chain“), 10 RDPs were devoted to devices (the and Computer Science, Alples d.d., “Intelligent appliances“ value chain), while the remaining 7 addressed CBD d.o.o., Cosylab d.d., Elgoline the value chain of “Intelligent home management“. d.o.o., INTECH-LES d.o.o., Kolektor group d.o.o., Lumar d.o.o., Robotina The overall goal was a transformation from an automated home to self-d.o.o., Roto d.o.o., Seltron d.o.o., learning adaptable home. In such advanced home, users’ behaviour is SI.mobil d.d., Strip's d.o.o., Špica d.o.o., followed through inbuilt intelligent appliances, and artificial intelligence Institute Jožef Štefan, TECES, TERMO-is utilized to mimic users’ habits and thus simplify home management TEHNIKA d.o.o., UP Università del and additionally reduce energy consumption. Litorale Andrej Marušič Institute, UL Biotechnical faculty, UL Faculty of Civil GeMMA participated in RDP5 – Intelligent planning of constructions of and Geodetic Engineering, UL Faculty buildings (TRL 3-4, completed in the beginning of 2018). 3D models of of Mechanical Engineering, UM Faculty several wooden-glass building modules were designed by UM FGPA and of Energy Technology, UM Faculty placed onto the tops of the models of selected real-world buildings in of Civil Engineering, Transportation Maribor, and various analyses were then performed. GeMMA’s role was Engineering and Architecture, UM Faculty to implement the activity “Insolation analysis and optimisation of solar of Chemistry and Chemical Engineering energy utilization” . in kemijsko tehnologijo, Lesarski grozd, wood industry cluster The first step was to integrate detailed models of a wooden-glass module and a building upgraded with the module into the wider real-world environment model in a form of 2.5D grid constructed upon the Additional information classified LiDAR point cloud. http://www.iq-home.si/en/ 39 GeMMA Activity Report 2016–2022 Selection of the basic topological element is among crucial decisions prior to running the simulation and analysis of insolation. 100 kWh/m2 974 kWh/m2 We have experimented on several geometric models (2.5D grid, voxels, triangle mesh), and finally accepted a hybrid model, where the wooden-glass module and the upgraded building are modelled by triangles while the environment was left in a form of 2.5D grid. We then adapted our photovoltaic potential estimation solution (pp. 35–37 in GeMMA 2000–2016 survey), but with an important difference that the original method only estimates suitability of buildings roofs for solar plants installation while the insolation analysis and optimisation performed in IQ HOME also require consideration of vertical surfaces, particularly those representing Figure 10: Direct solar irradiance on roof surfaces in the walls of the wooden-glass module and the upgraded building. 2.5D grid. (Source: own) Highly accurate estimation was achieved by considering the following factors: • Long-term measurements of the direct and diffuse irradiance (for the period 2004–2015 with temporal resolution 15 minutes) provided by ARSO were utilized to calculate the so- called Typical Meteorological Year – TMY; • Sun position simulation utilizes the algorithm published by the National Renewable Energy Laboratory (NREL), which computes the vertical and azimuth angle towards the Sun for a given micro-location at an arbitrary moment between years -2000 and 6000 with precision ±0.0003°; • Self-shadowing and shadowing from surrounding obstacles (e.g. buildings and terrain) utilizes the computed Sun position and several geometric attributes of the 3D model; • Shadowing from vegetation throughout the year uses satellite-based Leaf Area Index data; Figure 11: Solar irradiance on vertical walls and roofs in triangle mesh model. • Finally, the time-based integration of calculated irradiance (Source: own) for a given time period considers all the factors mentioned above. Figure 12: Solar irradiance on vertical walls and roofs in voxel model. (Source: own) 40 GeMMA Activity Report 2016–2022 Crowdtrust – An Integrated Trust Model for Crowd-sensing Systems in the Context of Smart Cities The project addresses the issue of involving residents in smart city systems using crowd-sensing systems. Crowd-sensing represents an important opportunity by actively involving residents (prosumers) Financed by in monitoring urban space and co-creating and improving urban infrastructure in a passive or active way. By integrating the population, EU (ERDF); MIZŠ RS we can significantly improve the quality of services, reduce the cost of investment in infrastructure and achieve measurable positive effects Duration in improving the quality of smart city services. Despite the obvious advantages, there are a number of unresolved challenges in the field 2017 to 2020 of crowd-sensing. The project focuses on two key ones. The first challenge is to ensure the highest possible level of trust in the data and information provided by people to the smart cities systesm. The ability Partners to separate real data from untrue data and to eliminate the factor of UM Faculty of Electrical Engineering and human subjectivity is crucial. To this end, the first emphasis is on the Computer Science, Inova IT d.o.o., development of an original method for the identification and treatment of false data and the subjectivity factor in data captured in crowdsensing smart city systems. The second challenge is the ability to convince residents that the data they provide is collected and processed exclusively for the purposes for which it is collected, that it is stored in a way that prevents abuse or misuse and that an adequate level of trust is ensured, which is an important prerequisite for the mass participation in crowd-sensing systems. The second focus of the project is on the development of innovative models and mechanisms for secure and confidential storage and processing of crowd-sensing data collected in a smart city environment based on block-chain technology and Ethereum platform, and the development of incentive mechanisms for participation in crowd-sensing environment. Both results are integrated into an integrated crowd-sensing trust model in the context of smart cities with support for two-way trust. 41 GeMMA Activity Report 2016–2022 PLACE – Predictive Analytics Based on Location-Associated Context Enrichment Programmes like Copernicus, GEOSS, and Galileo, together with the Internet of Things enabled devices, are daily generating huge amounts of Financed by sensory data that are becoming widely accessible. This has resulted in an exponential growth of captured data, which is becoming increasingly ARRS – Slovenian Research Agency more available and interoperable through open data initiatives (e.g. (contract J2–8176) the Digital Single Market of the EC), open source software (such as Duration GeoServer), and open standards (e.g. the OGC and INSPIRE Directives). These not only provide us with the opportunity to observe the natural 2017 to 2020 processes at high spatiotemporal resolution, but also enable us to monitor the causal relationships that are driving them. Although contemporary methods are capable of aligning geospatial data from Partners different information layers, assessing semantic features in a single or few aligned layers, and analysing raw data and/or extracted features, UM Faculty of Electrical Engineering and they often remain restricted to selected data types (e.g. hyperspectral Computer Science, Geodetic Institute of and LiDAR or sensors connected through wireless networks), where the Slovenia integration of domainspecific knowledge is relatively straightforward. Moreover, they do not incorporate data enrichment sufficiently, causing much of the data to be underexploited. As a result, many current state-of the-art methods are only capable of targeting situation assessment, while impact assessment is a characteristic of only a few domain specific decision support systems, where the contextual information is administered by users. Due to the complexity of interactions within and amongst natural processes and their relations to human activities, extraction of this so-called contextual information has only now become possible. In the PLACE (Predictive analytics based on Location-Associated Context Enrichment) project, we addressed contemporary challenges of structured context representation within the fusion of heterogeneous geospatial data sources and streams. For this purpose, we first acquired the context based on the development of temporal-spatial analyses using the concepts of computer geometry, topology and geospatial statistics, thus presenting new approaches to the recognition of contextual relationships and their structured presentation. Subsequently, we used context in predictive analyses, developing new methodologies for integrating structured contextual information to improve the accuracy of currently known methods of environmental simulations and regression models. The PLACE data fusion approach resembles a spiral model, capable of achieving “self-enrichment” of data during few iterations through the supporting processes. The latter include definition of geospatial entities (or objects), feature extraction, context structuring, and predictive analytics. 42 GeMMA Activity Report 2016–2022 Research activities were placed in two validation scenarios: 1. Prediction of microclimatic parameters, where we concentrated mainly to wind potential, heat load of buildings and air pollution. 2. Prediction of geomorphological changes, where we concentrated mainly to the study of glaciation or geomorphological changes due to the movement of frozen water – ice. The research results were published in prestigious scientific journals, where 17 papers were published in journals with SCI. Among them, 12 are in the A’ according to the Slovene research agency classifications, and 5 in A’’. In order to internationally promote the achievements of the project, 9 conference papers have been published. An patent was granted to us, too, that successfully passed a patent test in the United States. In this way, the research aims were outdone. Figure 13: Predictive analytics system arhitecture for achieving location associated context enrichment. (Source: own) 43 GeMMA Activity Report 2016–2022 WIBRANT – Wearable Integrated Smart Brace for Rehabilitation Monitoring and Diagnostic of Disorders in Muscular Functions WIBRANT addresses the “Healthy and active aging” priority subarea within the S4 priority area of “Health – medicine” . It was selected for Financed by funding within the ERDF and MIZŠ RS public call (2018) for proposals “to support Research and development projects (TRL 3-6)”. Its main goal EU (ERDF); MIZŠ RS is to develop a cutting-edge wearable sensory system in a form of an easy to wear smart flexible brace that enable improved tele-diagnostics Duration and tele-rehabilitation of muscular disorders, while also making these services affordable to elderly people. 2018 to 2021 The holistic solution provided by WIBRANT technology consist of the following innovative components that was realized in the Phase 1 (TRL Partners 3-4) of the project: UM Faculty of Electrical Engineering and Computer Science, Skylabs d.o.o., Inova • Carrier Brace Textile represents a comfortable wearable smart IT d.o.o., Institute for Sports Medicine at material, suitable for integration of the sensory system; UM Faculty of Medicine • Integrated flexible electronic sensory system (miniaturized Additional information microcontroller unit with memory and peripherals – sensors) perform energy-efficient patient monitoring in a real-world https://www.skylabs.si/wibrant/ environment and provide local data storage, data pre-processing capacities, and data transfer to mobile devices; • Mobile application provides efficient data transmission services interface between wearable electronics and its data storage on one side and the processing server. Besides this, it enables patients to monitor their rehabilitation progress; • Data analytics platform is data storage and management capacities that enable doctors to monitor the rehabilitation processes of their patients with dedicated advanced services for the assessment of musculoskeletal functions and treatment assignment. Furthermore, they get an in-depth view into the environmental parameters that influence the patient’s rehabilitation outside of the laboratory environment. 44 GeMMA Activity Report 2016–2022 Phase 2 (TRL 5-6) – here, the focus was on the completion of technology validation and technology demonstration in a relevant environment. For this purpose, a pilot environment was set-up at ISM. During these technology demonstration activities, involvement of potential customers (e.g. University Medical Centre Maribor) was conducted, while the first push towards integration into the global chain of values was made in partnership with Adidas, which strongly supports the project. Figure 14: The envisioned design and application of the Wibrant Brace (Source: WIBRANT project proposal). The WIBRANT consortium joined leading experts from diverse technological fields in order to successfully design and implement all these distinct but complementary subsystems. FERI MS was competent for smart textile design, while Skylabs has utilized its nanoelectronics expertise to develop the sensory system, and Inova IT was responsible for the mobile application and data transmission. GeMMA has used its expertise in spatiotemporal data analytics and provided advanced algorithms for fusion of sensory information with geospatial data (e.g. temperature, water moisture, and patient tracking data) and for machine learning algorithms for pattern recognition aimed to extract casual relationships between the sensory measurements and environmental data. 45 GeMMA Activity Report 2016–2022 After the completion of the project, the following main contributions of GeMMa were: • In addition to the consortium's support with existing data analysis techniques, GeMMA's researchers provided a new approach that transforms raw input data from Wibrant Smart Brace into geometrical deformations of the brace. Besides that, GeMMA also successfully introduced a new algorithm that maps deformations into a tensiomyography data; • A new method for visualization of geometrical deformations of the smart brace was created; • Simulation of muscle deformation with motion capture system was successfully conducted; • A new iterative approach to explainable feature learning was introduced; • Two journal articles listed below and one patent application which describes an apparatus and a process for real-time monitoring of deformation of smart elastic textiles based on measurements of electromagnetic characteristics were proposed. Figure 15: Wibrant mobile application for monitoring muscle activities. (Source: own) 46 GeMMA Activity Report 2016–2022 ION – Integration of Indoor and Outdoor Navigation ION addresses the “Mobility, transport and logistics” priority subarea within the S4 priority area of “Smart Cities and Technologies” . It was selected for funding within the ERDF and MIZŠ RS public call (2018) for Financed by proposals “to support Research and development projects (TRL 3-6)”. EU (ERDF); MIZŠ RS Its main goal is a personalized navigation system aimed to provide integrated indoor and outdoor routing services, particularly adjusted to Duration support people with mobility impairments. This objective was achieved in the first ION phase (TRL 3-4) by developing: 2018 to 2021 • A new software solution, providing generation of enriched navigation maps with mobility parameters and restrictions such Partners as stairways, width and height of sidewalks, routes’ steepness, UM Faculty of Electrical Engineering and location of elevators and ramps for individuals with mobility Computer Science, Inova IT d.o.o., Astron impairments by extending on the existing OpenStreetMap data d.o.o., UM Faculty of Civil Engineering, models and providing advanced spatial data editor with ability to Transportation Engineering and edit geometry and attributes directly in the spatial database; Architecture • A new indoor positioning system based on visual light communication (VLC) and compliant with mobile devices i.e. relying on existing mobile light detection sensors and existing Additional information infrastructure geometry descriptions; https://ion.inova.si/sl.html • Next generation navigation platform capable of mapping indoor and outdoor coordinates to the common coordinate system and switching between indoor and outdoor positioning, while considering personalized preferences for optimal path finding via dynamic path attribute filtering and rule evaluation engine. Within the second, pilot phase (TRL 5-6), the ION prototype platform was tested and validated in the selected real testing environments, set-up within bounds of the Municipality of Maribor and in one of the SPAR’s markets. GeMMA with its expertise in geospatial data processing algorithms and platforms development participated in ION by providing the back-end server-side infrastructure and services for navigation, enriched outdoor navigation map generation software, advanced graph-based path finding algorithm, collision detection, and indoor-outdoor coordinate matching. INOVA IT was responsible for frontend mobile applications, and ASTRON’s expertise in advanced electronics which they utilized for VLC electronics components, while FGPA's in-depth knowledge and understanding of urban mobility challenges were of indispensable value for technology steering, consolidation of user requirements and technology validation, and for communication with potential customers and end-user communities. 47 GeMMA Activity Report 2016–2022 Figure 16: Editing experience of the standardized spatial data with attribute schemas for easier information input. (Source: own) Figure 17: Visualization of the spatial graph streamed directly from the database which is used in pathfinding engine to provide detailed multipoint navigation for multiple transportation types with personalized rules. (Source: own) 48 GeMMA Activity Report 2016–2022 IPOT – An Integrated Pilot Environment for Sustainable Smart City Mobility GeMMA is participating in the project as an external contractor providing the technological backbone with integrated data fusion and back-office analytics for the following applications. The purpose of the iPOT project Financed by is to establish and integrate a next-generation mobility platform into the EU (ERDF); MGRT RS demonstration environment, which primarily, as a unique product on a global scale, enables the collection and processing of large amounts of data in real time. The project covers the field of mobility, transport, Duration logistics, the key goal of which is to increase the mobility of people and goods by providing reliable, flexible, accessible, safer, and green urban 2019 to 2023 and suburban services. The development of the modules is currently running in the direction of increased security in car parks, which aims to strengthen security in smart cities in both public and private sector. We Partners are also directly involved in the field of quality of urban living, the key goal UM Faculty of Electrical Engineering of which is to raise the quality of life in urban environments by ensuring and Computer Science, Iskra d.o.o., A1, sustainable green economic and social development. Comtrade, Globtel holding, Žejn group, Igea d.o.o., MSG Life, BAS5, Inova IT The key objectives of the project are: d.o.o., Spark d.o.o. • Establishment of a traffic control room; • Implementation of dynamic traffic regimes; Additional information • Smart parking; https://ipot.si/ • Integrated payment for mobility services. Due to its modularity, the iPOT solution is the foundation for effective global planning of other models of sustainable mobility in various urban agglomerations around the world. Figure 18: IPot car sharing control applications. (Source: own) 49 GeMMA Activity Report 2016–2022 Energy Management System for Electric Devices This was an ARRS basic research project lead by Laboratory for power engineering at UM FERI, where GEMMA was a collaborating research partner. The main goal of this project was to develop and validate a new Financed by concept of EMS. The developed energy management system eliminates ARRS – Slovenian Research Agency most of the mentioned shortcomings. In order to achieve the goal of the (contract J2–1742) project, the following tasks have been performed: Duration • To provide information that is essential for decision-making, 2019 to 2022 automated modelling of individual devices (generating units, storage tanks and energy consumers). The process was based on machine learning. It enabled automated creation and adaptation of Additional information device models based on measurements during normal operating states of devices. Developed device models were implemented https://ime.feri.um.si/energetika/ centrally on EMS or locally on switches / meters. When necessary, raziskovanje/projekti/sistem- the EMS was able to use the model to estimate the values of the za-upravljanje-z-energijo- observed variables at the time the action is performed before elektri%C4%8Dnih-naprav actually performing that action. These models supported the EMS in selecting the best measures. • A new EMS autonomous decision-making algorithm (AAO) was introduced to mimic market behaviour. The decisions of the AAO EMS were based on assessments of the condition of individual units for the production, storage and use of energy, which were provided by the models of these devices. Such an AAO EMS, based on an analysis of the situation and supply or demand of the aggregator, was able to accept or reject the offer, and at the same time was able to offer or demand services. The existing transmission of requests to users regarding the switching on and off of their devices was replaced by the exchange of supply and demand between AAO EMS and the aggregator. • An experimental EMS testing system was be set up to validate the proposed concept. It included controlled generation units (PV systems with micro-converters), energy storage (SCiB battery system and converter) and consumer devices (refrigerator, water heater, several air conditioners and other energy consumers). The experimental system was used to test the automated generation of device models, their ongoing adaptation, testing the properties of the developed EMS in parallel and island operation. It was also used to verify the suitability of the developed AAO EMS. 50 GeMMA Activity Report 2016–2022 Figure 19: Existing Experimental system at Laboratory for power engineering, UM FERI. (Source: https://ime.feri.um.si/energetika/raziskovanje/projekti/sistem-za-upravljanje-z-energijo-elektri%C4%8Dnih-naprav). 51 GeMMA Activity Report 2016–2022 LiDAR-facilitated Volunteered Geographic Information for Topographic Change Detection Acquisition of volunteered geographic information (VGI) or geographic crowdsourcing has gained increased attention from academia in the last decade, especially for topographic change detection, collaborative Financed by mapping and natural hazard monitoring. By means of VGI we can collect ARRS – Slovenian Research Agency positional data and georeferenced text, messages, photos or other (contract J2–1742) information, e.g. by tagging existing information with geographical location. Duration Different approaches can be used to motivate citizens and professionals 2019 to 2022 to participate in VGI, while the main motivation is usually a desire to cooperate in a worthy cause. Better data quality can be achieved if, together with laymen as data contributors, experts are cooperating. Aditional information https://ime.feri.um.si/energetika/ Topographic maps and data cover entire states. In the A4C4 quality raziskovanje/projekti/sistem- requirements scheme (Authority, Accuracy, Availability, Actuality, za-upravljanje-z-energijo- Completeness, Coverage, Consistency, Correctness), VGI wins in elektri%C4%8Dnih-naprav comparison with expert topographic data only in Actuality and conditionally in Correctness, but this is very significant for the Surveying and Mapping Authorities (SMAs). European SMAs renew topographic data periodically, e.g. once in every 3 years. In order to achieve high and geographically homogeneous Actuality of VGI input at any time (i.e. continuously), SMAs have to attract data contributors all over their countries. Therefore, the main goals of this research are: 1. To empower volunteers for easy and quick data collection; 2. To empower geodetic professionals to process these data with photogrammetric quality. A term facilitated volunteered geographic information was introduced, which describes the fact that the collection of VGI can be accelerated if the beneficial institution like SMA, supports volunteers e.g. with simple and user-friendly applications such as digital mapping interfaces or topographic data browsers. When VGI, especially volunteered photos are crossed with complementary georeferenced big data, e.g. with LiDAR point clouds or photogrammetrically derived digital surface models (LiDAR-like data) of whole countries or satellite images, new research directions emerge. Given the potentials offered by VGI and volunteered photos collection, the following three beyond state-of-the-art research problems arise: 52 GeMMA Activity Report 2016–2022 CENTRAL PROBLEM – How to optimize the methodology of topographic map updating to involve arbitrary VGI, single volunteered non-metric photos, LiDAR or LiDAR-like data and satellite images? CONTEXTUAL PROBLEM – How to support volunteers in the facilitated VGI, and in the volunteered non-metric photo collection for the purpose of full quality photogrammetric map updating: 1. For different topographical changes e.g. of road network, buildings, land use and land cover; 2. At different national topographic map scales of e.g. 1:5000 vs. 1:50.000. TECHNOLOGICAL PROBLEM – Specifically: 1. How to photogrammetrically orientate and georeference a single non-metric volunteered photo made by amateur camera or mobile phone; 2. how to extract and map 3D topographic changes from such an amateur photo only with the help of LiDAR or LiDAR-like data. Topographic map updating is usually done by photogrammetric survey, where imagery used to detect changes is professional, i.e. metric, orientated, stereo (in pair), and vertical (aerial - from an airplane). The main objective of the proposed research is the development of optimal methodology for a topographic map updating based on a mashup of volunteered geographic data, volunteered amateur photos and professional LiDAR or LiDAR-like data. We can summarize this with the following hypothesis: A topographic mashup of arbitrary VGI, volunteered photos, professional LiDAR or LiDAR-like data and/or satellite images can provide a professional standard quality input for 3D topographic change detection and mapping in the process of topographic map updating. 53 GeMMA Activity Report 2016–2022 EKOGEN – Economics of Farming with the Support of Geospatial Analyses EKOGEN addresses the issue of small farm economics by developing an advanced tool to support on-farm production planning. Because we are Financed by aware of the limited capabilities of small farms to reach out for advanced precision farming technologies, we limit our scope to the agricultural EU (ERDF); MGRT RS advisors, who we see as the bearers of digitization processes. Duration The developed products enable them to systematically plan production for their customers (farmers), and primarily consists of: 2020 to 2022 • Development of tools for visual modelling of production and definition of agricultural scenarios, which enables the integration of Partners open data sources for geospatial planning of crops and monitoring UM Faculty of Electrical Engineering and their development; Computer Science, FlawlessCode d.o.o., Igea d.o.o., ITC Murska Sobota • Upgrades to the existing expert system of economic calculations for the validation of specific scenarios, assessment of related costs and risks and potential sales revenues; Additional information • Adaptation of services for visual analysis and reporting, which https://flawless-code.com/ekogen/ enables the optimization of scenarios and generation of reports for third parties (banks, insurance companies, …) with methods of the artificial intelligence. The project was based on an existing market product, which is in use today by Slovenian consultants, which we further upgraded to a market product with a global reach. In the context of S4, the project addressed the priority area of “Sustainable Food Production” , within which it focused on the area of “Smart Process Planning and Process Control” . As an external contractor with the expertise on the geographic information systems our role was to develop advanced tools for geospatial data processing and establish development infrastructure that allows project partners to collaborate on the development process by integrating developed applications and services in the common environment. 54 GeMMA Activity Report 2016–2022 Figure 20: Development infrastructure for geoprocessing of the Sentinel-2 data and developed services. (Source: own) Figure 21: Visualization of the generated Sentinel-2 products. (Source: own) 55 GeMMA Activity Report 2016–2022 Implementing Digital Twins of Ecosystems of Agricultural Lands Digital twins have become a major technology trend and a critical component in the implementation of smart Financed by environments. With their ability to mimic the behaviour of real-world entities in virtual environments, they provide advanced monitoring, ARRS – Slovenian Research Agency diagnostics, prognostics, and optimization capacities. However, as their (contract L7–2633) implementation requires convergence of many technologies and non- Duration technical aspects, ranging from Internet of-things to artificial Intelligence with integrated domain-specific knowledge, their usage today is limited 2020 to 2023 to highly controlled environments, such as smart factories and smart homes. Their immense potentials to provide environmental intelligence, thus, remain unutilised, specially, when considering protection of Earth Partners from degradation through sustainable management of its natural UM Faculty of Electrical Engineering and resources and urgent actions against climate changes. Computer Science, Igea d.o.o., KGZS Murska Sobota Today, food production is amongst the main producers of greenhouse gases (GHG), while being under immense pressure due to the rapid urbanisation. It is, therefore, critical to address the trade-off between safeguarding food production, while lowering GHG emissions. This can only be achieved by deepening understanding of our interactions with agricultural ecosystems. The proposed project addresses contemporary challenges of digital twins for modelling such socio- environmental interactions by providing significant advances beyond state-of-the-art in the following aspects: • A new in-situ , capable of simultaneously capturing CO2, N2O and CH4 emissions, together with temperature and moisture of surroundings as well as levels of plant photosynthesis using quantum sensor with location data provided by Galileo; • A data harvesting system, intended for gathering and aligning IDEAL’s in-situ data with open Earth observation data sources (e.g. Copernicus satellite images, GEOSS thematic maps, and LiDAR data from Slovenian environmental agency) for common representation of spatiotemporal entities; • An advanced data fusion framework designed for mining IDEAL’s data sources by the principles of deep and feature learning for spatiotemporal extrapolations and crop-growth simulations; • Process optimization and visual analytics services for providing for prescriptive analytics capacities of socio-environmental interactions with the support of explainable artificial intelligence. 56 GeMMA Activity Report 2016–2022 As a result, IDEAL digital twin shall enable: • Farmers’ interaction with agricultural ecosystems; • Of green-house-gas emissions, soil health, and crop development parameters; • Of their changes during the time; • Optimization of farming processes, accordingly. In accordance with user-centric design, project development shall be governed by three complementary pilots, each addressing the specifics of a particular agricultural ecosystem that all together cover 98% of Slovenian farmland, namely, grasslands, arable lands, and permanent crops. Within each of the pilots, systematic data collections shall be conducted periodically during crop and grass growth, before and after all major farming activities, including tillage, fertilization, planting, and harvesting in order to ensure accurate profiling of the following parameters: • High-resolution GHG emission that includes carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4); • Soil health parameters and derived nutrition levels, as for example fertility indices, pH, and manganese; • Crop and grass development parameters based on their physical features like levels of photosynthesis productions and growth. In order to maximize the project potentials, IDEAL digital twin shall be plugged-in into existing precision farming infrastructure provided by industrial partner (namely Igea d.o.o.), turning natural ecosystem into a smart environment. IDEAL shall, thus, provide the necessary social innovation infrastructure to the researchers and practitioners that are currently struggling with low level of general digitalization in agricultural sector. 57 GeMMA Activity Report 2016–2022 Landscape Heterogeneity and the Forthcoming Agricultural Policy Measures in Slovenia The problem of biodiversity loss due to intensified agriculture and abandonment of high nature value farmland has been recognized as one Financed by of the major environmental problems in the European Union. Slovenia is trying to contribute to landscape biodiversity conservation through ARRS – Slovenian Research Agency, agricultural and other policy mechanisms, which include landscape MKGP RS (contract V4–2018) features as crucial biodiversity elements of its mosaic landscape. The aim of this project is to contribute to the re‐definition of agricultural Duration landscapes in Slovenia, and define the measures to improve the current situation with landscape features through the following goals: 2020 to 2022 • Identify areas for the conservation, restoration and establishment Partners of landscape features, and the definition of a set of landscape features suitable for the conservation of biodiversity in agriculture; UM Faculty of Electrical Engineering and Computer Science, UM Faculty of Arts, UM Faculty of Natural Sciences and • Prepare recommendations for the appropriate management Mathematics, Geodetic Institution Celje (conservation, restoration and establishment) of individual landscape features; • Prepare a classification and precise definition of landscape features relevant to both biodiversity and agriculture, and identify those landscape features that need to be maintained at the level of conditionality and those that should be maintained through climate and environmental schemes; • Define the dividing line between standard and above-standard measures, and prepare appropriate calculations for support for farmers; • Prepare the contents of mandatory and above-standard measures with a clear intervention logic, and propose an appropriate minimum share of agricultural area intended for non-production characteristics or landscape features; • Prepare starting points for determining the landscape features for inclusion among the eligible areas of income support under the direct payment scheme; • Develop an appropriate system for capturing data of individual types of landscape features in order to improve the databases; • Prepare an upgrade of the existing inventory of landscape features, adapted to the needs of the Ministry of Agriculture, Forestry and Food for the needs of preparation and effective implementation of the Strategic Plan from 2023 onwards, which will be suitable for inclusion in the Land Parcel Identification System (LPIS), and preparation of its maintenance proposal. 58 GeMMA Activity Report 2016–2022 The two main activities of GeMMA within the project are the development of spatial support information system for conservation of landscape features, and the testing of algorithm for landscape feature identification from available satellite, orthophoto, LiDAR and other spatial data for the selected area. Figure 22: Application for monitoring and planning of landscape features. (Source: own) 59 GeMMA Activity Report 2016–2022 Generalized Symmetries and Equivalences of Geometric Data An object has symmetry if there is a transformation (such as rotation, translation, scaling, reflection, etc.) that maps it onto itself. Being Financed by symmetric is a potentially very useful feature. Symmetries in the natural world have often inspired people to incorporate them when producing ARRS – Slovenian Research Agency tools, buildings, artwork etc. Therefore, it is important to be able to (contract N2–0181) detect symmetries in geometric data. Consequently, symmetry detection Duration became a challenging research topic particularly in pattern recognition, computer vision, computer graphics, and geometric modelling, where 2021 to 2023 it addresses problems such as object alignment, data compression, symmetrical editing, reconstruction of incomplete objects, or technical illustrations support. Partners UM Faculty of Electrical Engineering In the GeoSym project, three complementary groups of researchers and Computer Science, University of decided to join their efforts, knowledge and experience to address the West Bohemia (UWB) in Pilsen, Czech most current symmetry-related challenges: Republic • The computer graphics group from UWB: knowledge and ambitions in the development and implementation of geometric algorithms; • The group of mathematicians from UWB: study of formalized and generalized concepts in geometry and geometric algorithms; • The group from GeMMA at UM FERI: expert knowledge in EO data processing. Based on the previous research activities of all three groups and the identified symmetry-related challenges, the following research objectives were set: • Development of fast and reliable methods for detection of generalized symmetries, considering global, local, reflectional, rotational (axial), perfect and approximate symmetries, for common as well as highly non-uniformly distributed or perturbed input point sets and for continuous curves/surfaces; • Development of new methods for detection and computation of exact projective equivalences for finite sets of points (solutions of polynomial systems) and for further special algebraic varieties (mainly 3D surfaces), and of approximate equivalences and symmetries of perturbed objects; • Integration of symmetry detection into the methodology of semantic segmentation and object recognition in EO data in order to improve accuracy and enlarge the set of recognized classes, validated in a dedicated set of applications. 60 GeMMA Activity Report 2016–2022 Both participating universities will benefit from the cooperation. The Czech researchers receive a valuable feedback on how their existing and planned methods of detecting generalized symmetries will cope with the peculiarities of huge, noisy, incomplete, and unevenly distributed real data. On the other hand, the Slovenian researchers expect that the new symmetry-aware features will further improve their EO data fusion methodology, which already successfully classifies points of ground, buildings, and vegetation. Thus, some selected subclasses of buildings are expected to be identified, and the first steps towards the identification of tree species are also planned. a) b) Figure 23: a) Sentinel-2 satellite image of Northern Dalmatia with hinterland and b) best of the detected reflectional symmetries. (Source: own) In the first year of the project, GeMMA team provided a repertoire of EO datasets, representing separate buildings and trees extracted from pre-classified LiDAR, and the Czech side then performed tests of their own global reflectional symmetry detection algorithm. At the same time, GeMMA tested the same algorithm on EO raster (Sentinel-2) data. After that, the software framework has been developed to provide functionalities of reading EO data of diverse types, symmetry detection in this data, integration of detected symmetries into the feature extraction and data fusion, e.g. classification, segmentation, object recognition, as well as the visualization of input data and results. The GeMMA Fusion Suite software development kit (GFS SDK) was adapted for this purpose. Furthermore, GeMMA has also developed and implemented a couple of novel algorithms for global and local reflectional symmetry detection adapted to EO data, while the algorithm for rotational symmetry detection is about to be finished till the end of 2022. 61 GeMMA Activity Report 2016–2022 Figure 24: Local reflectional symmetry detected on a LiDAR point cloud of the Maribor Cathedral. (Source: own) Figure 25: Local reflectional symmetry detected on a voxelized LiDAR point cloud of Slomšek Square in Maribor. (Source: own) 62 GeMMA Activity Report 2016–2022 Development and Integration of New Data and Visual Analytics Algorithms into Investigation Platform Digital forensics is one of the cornerstones of information security today. The daily increase in fraud increases the need for more effective tools and approaches to ensure security. As part of the research and Financed by development project, we designed, developed, and integrated new ARRS – Slovenian Research Agency, algorithms for data and visual analytics into the existing analytical MNZ RS (contract V4–2117) platform of the client. In doing so, we improved digital forensic analysis of heterogeneous structured data, which represents a complex network Duration of different user domains (e.g. financial data flows, social networks, etc.). We also placed special emphasis on algorithms in the field of 2021 - 2022 artificial intelligence to perform improved data analytics, where hidden information and anomalies in complex networks could be found in an automated way. We upgraded visual analytics tool with the introduction of new visualization algorithms for the presentation of multidimensional data and the possibility of including new heterogeneous data flows. Together with the study of appropriate algorithms and their integration, we established cooperation between the security authorities of the Republic of Slovenia and the University of Maribor, Faculty of Electrical Engineering, Computer Science and Informatics (UM FERI) in terms of research challenges in national digital security. Figure 26: Components of the advanced investigation platform. (Source: own) 63 GeMMA Activity Report 2016–2022 Integration and Analysis of Heterogenous Data Streams in Investigation Platform Finding useful information in large amounts of heterogeneous data is an increasingly important part of digital forensics, with data in the Financed by form of complex networks most often helping to detect new patterns and acquire new knowledge. As part of the research and development ARRS – Slovenian Research Agency, project, we are designing upgrades to the existing analytical platform, MNZ RS (contract V2–2260) which enables the analysis of data in the form of a complex network. The solutions will enable the integration of new heterogeneous data streams Duration (e.g. social network data, telephone conversations, and air passenger data) into a graph format. The main innovation will be the possibility 2022 to 2023 of constructing a new graph over the input data, in case they are not sufficiently internally connected. We are additionally defining new data models, examining the suitability of existing algorithms over given data models, and developing new data and visual analytics algorithms to study new input data. As part of the project, we are also introducing the researchers to new advanced techniques. With this project, we are, thus, strengthening cooperation between the security authorities of the Republic of Slovenia and the University of Maribor, Faculty of Electrical Engineering, Computer Science and Informatics (UM FERI). 64 GeMMA Activity Report 2016–2022 An Integrated Approach for Conservation of Cultural Heritage Wall Paintings The preservation of cultural heritage (CH) is the only way to effectively transfer it to future generations. Given its importance, the key for a successful preservation of historical materials is a multidisciplinary Financed by approach to conservation issues. First step involves fundamental ARRS – Slovenian Research Agency understanding of an individual cultural heritage object and its condition (contract J2–4424) through knowledge-based learning and utilisation of advanced diagnostic and computational techniques. Second step requires Duration appropriate choice and implementation of conservation interventions needed, which can only be achieved through development of 2022 to 2025 methodologies, tools and materials to counteract, stop, and (ideally) revert the degradation process. However, for prolonged preservation of CH, the act of conservation itself is not regarded the final step; in order to Partners ensure the future of CH, validation of the effectiveness of conservation UM Faculty of Electrical Engineering interventions needs to be performed, as well as long-term monitoring of and Computer Science, Zavod za the CH itself. gradbeništvo Slovenije, Javni zavod Republike Slovenije za varstvo kulturne Wall paintings represent one of the most important types of cultural dediščine, Igea d.o.o. heritage. As an integral part of architecture, their state of preservation usually reflects the history of architecture itself by displaying degradation, damage, numerous historical treatments and redesigns. Wall paintings, embellishing architectural façades are particularly prone to decay since they suffer direct exposure to environmental conditions. Moreover, digital documentation can significantly improve the understanding of their present state, as well as planning of preservation maintenance, presentation and promotion. Advanced techniques, such as LiDAR, can be particularly useful in providing an overall assessment of the entire surface investigated, which can be profitably used to identify those specific areas in which further analytical measurements, sampling, laboratory analysis or conservation-restoration treatments are required. Sometimes sites of interest are difficult to access, or the test fields where conservation and restoration interventions have been carried are no longer available after scaffolding has been removed’. The use of advanced equipment such as drones is therefore highly desirable, since they offer faster, more advanced (comparison of 3D models taken before and after the procedure), safer and cheaper analysis (no “roadblocks”, scaffolding, permits, or safety requirements e.g. helmets and seat belts) of the object condition. Furthermore, when new materials are developed (such as the new cleaning and consolidation procedures presented in this project), it is very important to monitor the condition of the materials following such interventions, as the long-term effectiveness of such interventions is often still unexplored. 65 GeMMA Activity Report 2016–2022 Patient Individualised Management of Endometrial Cancer Gynaecological cancers represent a unique group of cancers associated with the endocrine physiological regulations in the body. Standard Financed by management of these cancers often has a significant impact on the hormonal balance in women and can lead to significant debilitating ARRS – Slovenian Research Agency consequences due to early menopause or loss of reproductive function. (contract J3–4523) Endometrial cancer is the most common gynaecological malignancy Duration in the developed world and in women younger than 40 years represent up to 5% of cases and around 20% of women are diagnosed before 2022 to 2025 menopause. Although most endometrial cancers are diagnosed early, up to 20% progress to high-stage carcinoma. Current diagnostic approaches fail to Partners identify high-risk disease that is apparently early stage at presentation. UM Faculty of Electrical Engineering and This indicates the need for improvement in risk assessment and Computer Science, University medical subsequent management of these women. center Maribor, UM Faculty of Medicine, UM Faculty of Chemistry and Chemical Current risk assessment is based on clinical or integrated molecular Engineering, University medical center group classifications endorsed by the ESGO/ESTRO/ESP guidelines. Ljubljana These classify endometrial cancer into 4 distinct groups. These groups are POLEmut (Polymerase Epsilon–Mutated), MMRd (Mismatch Repair Deficiency), p53abn (p53 Abnormal) and NSMP (no specific mutational profile). The NSMP represents the largest group. Considering the heterogeneity in prognosis, there is a great need for additional specific biomarkers. Improved risk assessment will enable therapy de-escalation and a safer approach to non-standard, fertility sparing therapy (FST). This will ultimately enable individualised counselling and patient focused treatment. Following this path, we should be able to shift the focus from oncological outcomes to improvement of long-term patient reported outcomes (PROs). In the project, we will address the current unmet needs in women with endometrial cancer by: • Identifying new biomarkers (WP1) to improve risk stratification, de-escalating therapy, identifying candidates for non-standard therapy, such as FST or hormone replacement therapy; • Developing conventional and smart risk stratification algorithms (WP3) to incorporate these biomarkers; • Developing minimally invasive methods of diagnostics and screening (WP2) that would allow accurate risk stratification, early diagnostics and possible screening in high-risk populations. Finally, following the results of our research, our ultimate goal is to improve PROs (WP4). We will first recruit patients at both national tertiary centres to obtain the necessary biological samples and precise tumour imaging data. Through sample analysis, we will determine the established molecular classification and analyse for the presence of new biomarkers. 66 GeMMA Activity Report 2016–2022 In addition to evaluating biomarkers in standard therapy, UMC Maribor will lead research of the molecular classification and biomarkers role in FST. This will provide fresh insight on the impact of tumour biology on reproductive and oncological outcomes of FST. Furthermore, we will focus on the possibility of obtaining the diagnosis and the biomarker-based risk assessment non-invasively. The project will focus on developing liquid biopsy methods and analysis of cell-free DNA and cell-free RNA in women with endometrial cancer to enhance individualised management. The main purpose of introducing novel biomarkers to clinical practice is to improve patient tailored management and possibly use less aggressive management in low risk patients. Hence, we have designed “in-vitro” studies of standard and unconventional therapeutic approaches to molecularly characterised endometrial cancer. For this purpose, we will for the first time characterise our own and commercially available endometrial cancer cell lines. The findings of these studies will have major implications for the design of subsequent clinical trials. All the knowledge gained through our project will be integrated to design a better, clinically applicable risk stratification model. The findings will culminate in better possibilities for tailored management and precision medicine, especially in young, low-risk women with endometrial cancer. 67 GeMMA Activity Report 2016–2022 DIGISAD –Development and introduction of digital tools to support fruit production Fruit production in Slovenia and beyond is becoming an extremely demanding industry due to the need to adapt to climate change, increasing Financed by environmental requirements, and specific market requirements, which both traders and consumers form. To help fruit growers adjust to ARRS – Slovenian Research Agency, changed conditions more easily, we will develop tools that will help them MKGP RS (contract V4–2230) make decisions regarding the approach to producing the highest quality fruit while constantly searching for "internal reserves" and optimizing Duration production processes to achieve economical production. The developed 2022 to 2025 digital tools will enable easy and timely access data from meteorological stations. Digital tools will thus include a digital handbook with guidelines for identifying pests and diseases, early crop predictions based on fruit Partners images, and a tool for determining critical points and the level of risk when transitioning to more demanding production systems (e.g. from UM Faculty of Electrical Engineering conventional or IP in EKO). As part of the project, we will also continue and Computer Science, UM Faculty of developing a customized sprayer. Chemistry and Chemical Engineering, Kmetijski inštitut Slovenije, Kmetijsko gozdarski zavod Maribor, Kmetijsko gozdarski zavod Nova Gorica, UL Faculty of Mechanical Engineering 68 GeMMA Activity Report 2016–2022 COMPROMISE – Data Compression Paradigm Based on Omitting Self-evident Information Data compression is one of the traditional disciplines of computer science, but one that has made no significant progress in recent decades. It has also failed to keep up with new scientific trends, where Financed by new devices collect ever-increasing amounts of highly heterogeneous ARRS – Slovenian Research Agency data. These data are compressed using either domain-dependent or (contract J2–4458) general-purpose methods. The general-purpose methods are well- known lossless solutions from 30 years ago (e.g. RAR or ZIP). They Duration achieve generality by handling the data stream on the level of bytes, ignoring potential higher-level relations in the data. Domain-dependent 2022 to 2025 methods are lossy, near lossless, or lossless. Lossy methods operate by transforming the data into frequency space, performing the quantization there, and encoding the remaining values in a lossless manner, whereby Partners the lossless part is typically domain-dependent as well. Near lossless UM Faculty of Electrical Engineering and and lossless methods are significantly different and typically prediction Computer Science, University of West based. However, the prediction is made from a narrow spatial and/ Bohemia (UWB) in Pilsen, Czech Republic or temporal context, which reduces its efficiency. Most methods are symmetric, which means that decoding is performed by the same pipeline as encoding, only in a reversed order. The disadvantage is that the time complexity of decoding is the same as that of encoding, which requires similar infrastructure for both the encoder and the decoder. Finally, each type of data requires a specific solution that is not transferable to other types of data (e. g. audio compression is completely different from compression of raster images). In the COMPROMISE project, we aim to develop a new data compression methodology which will be largely domain-independent and asymmetric. By using a unified pipeline of procedures, the methodology will be suitable for lossy, near lossless, and lossless compression. Domain independence will be achieved by forming feature repertoires in different domains and linking them to a unified domain-independent taxonomy. In our case, a feature will be any piece of information with high discriminative or predictive value for human interpretation or machine processing (e.g. computer vision, classification) of a data stream. The obtained repertoire of features will be reduced through a domain-independent iterative optimisation process, as long as the set of remaining features will allow the restoration techniques to perform satisfactory reconstruction of the input data. The compression pipeline will be the same for lossy, lossless, and near lossless compression, except that the output in the latter two cases will include the residuals, obtained as the difference between the original and the restored data. The data decompression will be much simpler and will consist of features and residuals decoding, restoration of data from features, and applying residuals in cases of lossless or near lossless mode. This will set the requirements for the decoder substantially lower than those for the encoder. The concept of domain-independent features also allows the information about higher-level relations in the data to be preserved in the compressed form, which improves the reusability of data on different semantic levels. 69 GeMMA Activity Report 2016–2022 In order to demonstrate the universality and domain independence of the methodology we will use raster images, digital audio, biomedical signals, and sparse voxel grids in our study. These domains differ in both the data dimensionality and dynamism, while addressing two human perceptual systems – vision and hearing. The proposed domain independent methodology will be implemented with a unified platform, which will be used to demonstrate the efficiency and universality of the COMPROMISE methodology, to validate the key performance indicators, and to verify the scientific hypothesis. By using the methodology, we expect to achieve better lossless and near lossless compression ratios than existing domain-dependent methods, which will set the foundation for a new generation of data compression methods. 70 GeMMA Activity Report 2016–2022 International Projects 71 GeMMA Activity Report 2016–2022 HOLISTIC – Wildfire Monitoring and Management System HOLISTIC aims at development of comprehensive wildfire monitoring and management system at the Adriatic seacoast. GeMMA provided a Financed by group of experts for environmental and Earth observation data processing for the Municipality of Ajdovščina, where the system is being evaluated European Comission, IPA Adriatic Crossin operational environment. An advanced GIS has been developed that Border Cooperation Programme provides real-time decision support to fire fighters, civil protection, and other first responders. This GIS allows for integration of real-time video-streams from thermal cameras and supporting information acquired by Duration drones. Integrated analytics tools include navigation support for rescue teams as well as information support for evaluation of burned areas. 2014 to 2016 In addition, the system integrates tracking of units for their improved coordination, supported by automatic routing and mapping of obstacles. Partners UM Faculty of Electrical Engineering and Computer Science, Municipality of Ajdovščina; DAT-CON d.o.o; Slovenia Forest Service; 20 partners from Croatia, Bosnia and Herzegovina, Montenegro, Serbia, Albania, Greece and Italy Additional information https://www.adriaholistic.eu/ Figure 27: Current thermal camera view (green), densely populated areas (red), and the route of the fire brigade (blue) from its current position (A) to the fire location (B), together with all the layers of critical infrastructure. (Source: own) 72 GeMMA Activity Report 2016–2022 MAHEPA – Modular Approach to Hybrid- Electric Propulsion Architecture MAHEPA is a research project aimed to boost research in the field of low emission propulsion technology. Its mission is to open up the potentiality for the series production of greener airplanes in order to Financed by support European environmental goals in aviation, which require a 70% EU (H2020 Programme) reduction of greenhouse gases until 2050. MAHEPA developed new components in a modular way to power two Duration four-passenger hybrid electric airplanes that flew in 2020 and 2021, respectively. The first was equipped with a hybrid powertrain utilizing 2017 to 2021 an internal combustion engine, and the second was a fuel cell hybrid-powered aircraft, showcasing the possibilities for zero-emission long-Partners distance flight as a concrete example of this innovative propulsion technology. UM Faculty of Electrical Engineering and Computer Science, Pipistrel Vertical The main results of MAHEPA project were thus novel, modular, and Solutions d.o.o., Compact Dynamics scalable hybrid-electric powertrains capable of running on alternative GmbH, DLR, University of Ulm, H2FLY fuels or hydrogen with zero emissions. However, not only new GmbH, Delft University of Technology, technologies has been developed, but also extended studies have been Politecnico di Milano, UM Faculty of Civil made on regulatory implications, airport infrastructure requirements, Engineering, Transportation Engineering airspace procedural practices, operational safety, operating costs and and Architecture emission models resulting in a unique outlook for regulators, aviation industry, operators and potential investors. Additional information https://mahepa.eu/ The role of GeMMA was to provide an attractive digital presentation of the designed airplanes with an emphasis on their propulsion systems. We were thus developing: 1) An augmented reality presentation of the airplanes’ exterior parts and propulsion systems by utilizing the HoloLens AR technology; 2) A realistic 3D presentation of the aircraft cockpit and cabin by utilizing HTC Vive virtual reality headset; 3) Multimedia presentation of the airplanes’ technical data and generally about the MAHEPA project, displayed on a smartphone and interactively controlled through quick response (QR) codes on a physical model of one of the two developed aircrafts. 73 GeMMA Activity Report 2016–2022 Figure 28: Physical model of the MAHEPA aircraft and demonstration of the MS HoloLens AR technology utilization (Foto: D. Podgorelec). Figure 29: MAHEPA showroom on AERO 2019 general aviation fair in Friedrichshafen, Germany, powered by GeMMA by digital exhibition tools developed by GeMMA (Foto M. Marksel). 74 GeMMA Activity Report 2016–2022 SmartVillages – Smart Digital Transformation of Villages in the Alpine Space Alpine Space rural communities are deprived of highly needed jobs, good provision of services as well as a favorable climate for enterprenuership and social innovation, which result in a brain drain. Digitalization is a Financed by promising approach to counter the situation. A Smart Village approach Interreg Alpine Space for mountain areas could unlock the potential of local actors to make their region a more attractive place to live and work. In cooperation with thirteen partners Smart Villages aims to brings together policymakers, Duration business, academia, and civil society in a quadruple helix approach to 2018 to 2021 improve the framework for innovation through new forms of stakeholder involvement facilitated by Information and Communication Technologies. The Project is a strategic initiative of EUSALP Action Group 5 and follows Partners an integrative, participatory approach implying a city - village dialogue. UM Faculty of Electrical Engineering The project contributes to better framework conditions for innovation on and Computer Science, Swiss Centre two aspects: the organisational and societal part - working with regional for mountain regions Schweizerische support groups involving policy level, academia, business, and civil Arbeitsgemeinschaft für die Berggebiete, society - and the technical part - DEP (Digital Exchange Platform) and University of Ljubljana, SmartiS City Toolbox with new digital products - and combining the strengths of both d.o.o., Poliedra – Politecnico di Milano, sides. Finally, the transfer of the results to the policy level contributes to Agenzia di Sviluppo Gal Genovese, improve the political framework conditions for digital innovation. Energie und Umweltagentur Betriebs- GmbH Niederösterreich, Association The DEP enables the transnational knowledge sharing of the overall pour le Développement en REseau project findings and European Smart Villages best practices between des Territoires et des Services, project partners and the wider public. Toolbox is a combination of tools, Regionalverband Südlicher Oberrhein, methods and techniques that provide the main ingredients of a smart Bodensee Standort Marketing GmbH, village. It supports and guides project partners, but also other regions Tiroler Zukunftsstiftung, Software within Alpine Space through the participatory establishment of a smart Competence Center Hagenberg, Region village environments in their regions. Final version of the DEP is available Luzern West at https://smart-villages.eu/language/en/home/. Additional information Our role was to provide services for smartness assessment, integration of the partners, services for matchmaking and toolbox methods based https://www.alpine-space.org/projects/ on survey input, survey management and good practices with automated smartvillages/en/home language translation via Google Translation APIs: Services are consumed by the DEP itself and provide public endpoints for data export to other project partners for further data analytics following the OpenAPI specifications. 75 GeMMA Activity Report 2016–2022 SMART2 – Advanced Integrated Obstacle and Track Intrusion Detection System for Smart Automation of Rail Transport SMART2 project was aimed to build a holistic trackside obstacle detection (OD) and track intrusion detection (TID) systems with correspond-Financed by ing interfaces to a central decision support system (DSS). Different remote sensing technologies, including video imaging cameras, thermal EU (H2020 cascade funding Shift2Rail) imaging cameras, 3D time of flight cameras, radar, and LiDAR, were thus incorporated by individual partners of the strong internation consortium. FOKUS TECH d.o.o. and GEMMA as their subcontractor were responsible Duration for 3D LiDAR sensors for monitoring dangerous areas at level crossings 2019 - 2022 of railway and roads. FOKUS TECH has developed its own 3D LiDAR, which is mounted on a pole near a level crossing, while GeMMA has developed software for detecting obstacles on the crossing near and Partners within the dangerous area. The former are indicated with green bounding boxes and the latter with the red ones in the service application, while the UM Faculty of Electrical Engineering and dangerous area is bounded with blue polygon edges. The sensor covers Computer Science; Universität Bremen, a spatial angle of 60° x 30° and has a range of 40 m. This is sufficient for Germany; OHB Digital Services GmbH, most level crossings on double-track lines. However, two systems could Germany; Univezitet u Nišu, Serbia; be used for larger crossings and more complex situations. HARDER Digital SOVA d.o.o. Niš, Serbia; Universitatea Tehnica Cluj-Napoca, Romania; Newcastle University, The United Kingdom; FOKUS TECH napredne tehnologije d.o.o., Slovenia Additional information https://smart2rail-project.net Figure 30: LiDAR sensor (in a yellow circle) installed on a pole at the level crossing (Foto: archive of FOKUS TECH d.o.o.). 76 GeMMA Activity Report 2016–2022 Figure 31: Test scenario and acquired point cloud with obstacles (Foto: archive of FOKUS TECH d.o.o.). Figure 32: Detected human (red bounding box) in the dangerous area (blue polygon). (Source: own) 77 GeMMA Activity Report 2016–2022 STAMINA – Demonstration of Intelligent Decision Support for Pandemic Crisis Prediction and Management Within and Across European Borders STAMINA develops an intelligent decision support toolset for pandemic prediction and management and demonstrates its use by practitioners at Financed by national and regional levels within and across EU borders. The STAMINA toolset enables national planners and first responders to anticipate and EU (H2020 Programme) respond to the the “known-unknowns” in their daily effort to enhance health security. Main functionality of the toolset includes: Duration • Real-time web and social media analytics aiming at public trust 2020 to 2022 monitoring and flagging possible disease outbreaks; • POCT (point of care testing) and smart wearable diagnostic devices Partners for first line screening; UM Faculty of Electrical Engineering • Predictive modeling of pandemic outbreak and its impact, and Computer Science, EXUS, ICCS, along with decision-making support in implementing mitigation AIT GmbH, Crisisplan B.V., Intrasoft strategies; S.A, Squaredev, Satways, Trilateral Research LTD, EE Viopliroforikis • Early Warning System; Kai Ypologistikon Epistimon, Eigen vermogen van het Instituut voor • Crisis management tool defining the roles and actions of key actors Landbouw- en visserijonderzoek, MCS during crisis management; Datalabs, Innosystems, Brunel University London, Istituto per L'interscambio • Scenario Generation tool for creation of training scenarios; Scientifico, Verisk Analytics GmbH, Westfaelische Wilhelms-universitaet • Common Operational Picture as the main interface of the solution Muenster, BYS group, Technologicka enabling timely and coordinated response. Platforma Energetickabezpecnost CR ZS, Institut Pasteur de Tunis, Beia Consult The toolset is accompanied by a set of Guidelines on effective International SRL, Erasmus Universitair implementation of risk communication principles and best practices in Medisch Centrum Rotterdam, cross-organisational preparedness and response plans. The use of the STAMINA toolset will be demonstrated through 12 national and regional small-scale demonstrators and one large-scale cross-border simulation exercise involving all consortium partners. 78 GeMMA Activity Report 2016–2022 OBJECTIVES METHOD The STAMINA vision has been STAMINA will be developed using a combination of pre-existing technology designed through a user perspective not currently used by health emergency with 5 main objectives: planners or first responders in Europe in 1 their daily practice of pandemics Smart support management. It will function through its partnership with a set of guidelines and Create a set of guidelines platform for best practices, and a key focus on ethics and best practices to and public trust. improve preparedness pandemic 5 and response. The method involves gathering data in order to predict potential threats, assess prediction and Ensure the sustainability impact on financial and societal levels, of the STAMINA solution. and recommend mitigation actions. An management intelligent decision support toolset with a map-based interface will be the main technical outcome of the project. ABOUT 2 Infectious diseases have the potential to THE STAMINA TOOLSET result in serious cross-border public 4 health threats. Management of this type Provide stakeholders with Real-time web and social media of crisis remains a serious challenge due novel, easy-to-use analytics to number of people involved, the Improve cooperation software tools that different legal, administrative, Wearable diagnostic devices between and within the EU complement EU-level professional and political cultures, and Member States and systems. the lack of transboundary crisis Predictive modelling neighbouring countries. management infrastructures. 3 An early warning system STAMINA helps to overcome these challenges by providing improved Increase diagnostic A crisis management tool decision-making technology to national capability. planners, pandemic crisis management WHO ARE WE? A scenario generation tool practitioners and first responders at a regional, national and European level. The STAMINA consortium is made up of 37 A common operation picture The toolset will be accompanied with a organisations from across the EU (and clear guide to how it can be used in line beyond) who have united to deliver a with international standards and two-year Horizon 2020 innovation project. legislation. Our project is led by: Decisionmakers, policymakers, national Visit: www.stamina-project.eu planners and public authorities Email: info@stamina-project.eu Health care workers, regional emergency management agencies, first responders and NGOs Social scientists, (bio)informaticians, This project has received funding from research organisations, IT experts the European Union’s Horizon 2020 research and innovation programme under grant agreement No 883441 Figure 33: Project abstract. (Source: https://stamina-project.eu/flyer-and-brochure/) There are three laboratories involved on behalf of UM FERI in this project: • Laboratory for Geospatial Modelling, Multimedia and Artificial Intelligence (GeMMA); Partners • Laboratory for System Design; Assistance Publique - Hopitaux de • System Software Laboratory Paris, Zdravstveni dom dr. Adolfa Drolca Maribor, Johanniter Osterreich ausbildung GeMMA coordinated the developments in support of technology und forschung gemeinnutzige GmbH, integration and implementation of real-world use-cases, carried out at Cruz Roja Espanola, Red Cross district Zdravstveni dom dr. Adolfa Drolca Maribor. 5 Bucharest, Ethniko kai kapodistriako panepistimio athinon, Fundacion de la co- munidad Valenciana para la investigacion, promocion y estudios comerciales de Valenciaport, Ayuntamiento de Valencia, ethnikos organismos dhmosias ygeias, Nacionalinis visuomenes sveikatos cen- tras prie sveikatos apsaugos ministerijos, Ministry of health, Turkiye Cumhuriyeti Saglik Bakanligi, Observatoire National des Maladies Nouvelles et Emergentes, Department of health, ministry of the interior of the Czech Republic, Institutul de Virusologie Stefan S. Nicolau. Additional information https://stamina-project.eu/ 79 GeMMA Activity Report 2016–2022 CobotSense – Intelligent 3D Safety Sensor for Cobot Applications COVR (Being safe around collaborative and versatile robotos in shared spaces) is an EU-funded H2020 project aimed to determine protocols how to test and validate safety for collaborative robot (cobot) applications or Financed by components. The protocols are being developed by third parties in series COVR award (H2020 cascade funding) of smaller projects – COVR awards, CobotSense being one of them. In the Factories of the Future, humans and cobots will share common Duration workspace to enable more flexible and cost-effective production. In this new paradigm, small cobots are already a reality, but this is not the 2020 to 2021 case for big robots and heavy-duty applications. New 3D safety sensors and intelligent control systems (ICS) are needed for these new cobot Partners applications, but they are unfortunately not available on the market yet. To fill this gap, the CobotSense partners have developed such a novel UM Faculty of Electrical Engineering and laser-based 3D safety sensor with an associated ICS. Computer Science, FOKUS TECH d.o.o., FANUC ADRIA d.o.o. The goals of CobotSense, all successfully achieved, were: • Advances on development of the prototype 3D LiDAR sensor for Additional information cobot applications. The key performance indicator (KPI) here is reaching the scanning speed of up to 5 frames per second; https://www.safearoundrobots.com/ home • ICS development. The KPI here is the validated protective separation distance (PSD) calculation for speed and separation monitoring (SSM) cobot applications. This was achieved by developing and integrating the software modules for: registration of the scanned point cloud and the robot’s geometric data; determination of the robot’s pose by the developed forward kinematics model; the scene segmentation into the robot, static obstacles, and operator; motion prediction for both, robot and operator; the PSD calculation; and real-time adjustment of the robot’s speed in order to realize the SSM principle; • Specification of use cases for testing the integration of a robot, operator, sensor and ICS in laboratory and industrial environments. A COVR case story was described (https://youtu.be/cEMr60nl1hE), and the COVR protocol for testing SSM cobot applications monitored with 3D sensors was developed and specified (https://covrfilestorage.blob.core.windows.net/documents/protocols/ ROB-MSD-3-Test_3D_Safety_Sensors_in_Speed_and_Separation_Monitoring_Cobot_Applications.pdf). Two UM FERI labs (Laboratory for Geospatial Modelling, Multimedia and Artificial Intelligence, and Laboratory for Cognitive Systems in Mechatronics) participated mainly in the ICS development and COVR protocol preparation. 80 GeMMA Activity Report 2016–2022 Figure 34: Cobot (yellow), operator (magenta), static obstacles (green), and noise (red) in the voxelized scene. (Source: own) Figure 35: Bounding boxes of the robot links and the operator. (Source: own) 81 GeMMA Activity Report 2016–2022 InspectRAIL – Autonomous Mobile Robots for Inspection of Railway Lines Railway safety is threatened by landslides, falling rocks and trees, floods, collapses of rail lines and load-bearing structures, torrents, and Financed by other hazards. Unlike the SMART2 project, where statically installed LiDAR sensors were used to monitor dangerous areas at level crossings EU (H2020 funded RIMA Network) of railway and roads, much wider open sections of railways require a new concept for inspection. InspectRAIL thus introduced inspection Duration by autonomous mobile robots, which move through dangerous places along the railway lines on steel cables stretched between the catenary 2022 masts and operate in all weather conditions. The robots are equipped with LiDAR sensors and video cameras which inspect the railway tracks. Partners The acquired data is automatically processed and sent to the Intelligent Control System (ICS), signaling system, traffic control, and maintenance UM Faculty of Electrical Engineering and centre for further actions. As movement requires higher frame rates than Computer Science, FOKUS TECH d.o.o., statically mounted sensors, a contemporary Ouster OS1 LiDAR sensor ALTPRO d.o.o., Croatia was chosen. GeMMA participated as a subcontractor of the project partner FOKUS Additional information TECH d.o.o. and was responsible for developing the ICS, where several challenges had to be addressed. Detection of rails was initially designed https://rimanetwork.eu/ on the reflectivity measurements performed by the utilized sensor, but this solution was not stable enough in varying weather conditions, so we had to rely entirely on geometric data only. Figure 36: The Camera view of the ICS with alert indicator due to detected obstacles. (Source: own) 82 GeMMA Activity Report 2016–2022 In normal conditions, the rails are about 15 cm above their surroundings, but we had first to align the point cloud due to sagging of the wires and the robot tilt due to wind. An external inertial measuring unit (IMU) was used for this task, which requires synchronization with LiDAR among all. The obstacle detection is then performed in a simple manner. LiDAR points detected below the top of the sleeper (TOS) plane in the section of the railway track correspond to lack of the ballast and must be reported as a potential danger. The second dangerous section is above the top of rail (TOR), where no obstacle should be detected. Besides detecting the obstacles and sending alerts, ICS must also display the mileage which represents the position of the robot relative to the reference rail system. The mileage is calculated from the data of the encoders of the drive motors and the radio frequency identification (RFID) tags on the poles of the overhead contact line. Figure 37: Besides the obstacles above the rails, the system Figure 38: The LiDAR view of the ICS with detected rails must also detect holes in the track ballast (Foto: archive of (yellow) and obstacles (red) in the point cloud (Foto: archive of FOKUS TECH d.o.o.). FOKUS TECH d.o.o.). Figure 39: The robot (in the yellow circle) with LiDAR and camera during the detection of obstacles in the test area (Foto: archive of FOKUS TECH d.o.o.). 83 GeMMA Activity Report 2016–2022 PLOTO – Deployment and Assessment of Predictive Modelling, Environmentally Sustainable and Emerging Digital Technologies and Tools for Improving the Resilience of IWW Against Climate Change and Other Extremes PLOTO aims at increasing the resilience of the Inland WaterWays (IWW) infrastructures and the connected land- infrastructures, thus, ensuring reliable network availability under unfavourable conditions, such Financed by as extreme weather, accidents and other kind of hazards. EU (Horizon Europe Programme) Our main target is to combine downscaled climate change scenarios (applied to IWW infrastructures) with simulation tools and actual Duration data, so as to provide the relevant authorities and their operators with an integrated tool able to support more effective management 2022 to 2026 of their infrastructures at strategic and operational levels. Towards this direction, PLOTO aims to: Partners UM Faculty of Electrical Engineering and • Use high resolution modelling data for the determination and the Computer Science, Intrasoft International assessment of the climatic risk of the selected transport infrastruc-SA, Exus, Budapesti Muszaki es tures and associated expected damages; Gazdasagtudomanyi Egyetem, Diadikasia • Use existing data from various sources with new types of sen- Business Consulting Symvouloi sor-generated data (computer vision) to feed the used simulator; Orgszagos Egyesulet, Universite de • Utilize tailored weather forecasts (combining seamlessly all avail-Liege, Asministrativa Fluviala a Dunarii able data sources) for specific hot-spots, providing early warnings de Jos R.A. Galati, Universitatea Danubis with corresponding impact assessment in real time; din Galati, Romanian River Transport • Develop improved multi-temporal, multi-sensor UAV- and satel- Cluster, MAV Magyar Allamvasutak lite-based observations with robust spectral analysis, computer vi-Zartkoruen Mukodo Reszveny Tarsasag, sion and machine learning-based assessment for diverse transport National Technical University of Athens, infrastructures; 84 GeMMA Activity Report 2016–2022 • Design and implement an integrated Resilience Assessment Plat- form environment as an innovative planning tool that will permit a quantitative resilience assessment through an end-to-end simula-Partners tion environment, running “what-if” impact/risk/resilience assess- RISA Sicherheitsanalysen GmbH, Ilmati- ment scenarios. The effects of adaptation measures can be inves- eteen Laitos, Budapesti Szabadkikoto tigated by changing the hazard, exposure and vulnerability input Logistzikai Zrt., Societal and Resilience parameters; Climate Change Centre of Excellence, • Design and implement a Common Operational Picture, including Service Public de Wallonie, Aristotelio an enhanced visualisation interface and an Incident Management Panepistimio Thessalonikis, European System. road transport Telematics implementa- tion coordination organisation - Intel- The PLOTO integrated platform and its tools will be validated in three ligent transport systems & services case studies in Belgium, Romania and Hungary. Europe, Satways 85 GeMMA Activity Report 2016–2022 PrAEctiCe – Potentials of Agroecological Practices in East Africa with a Focus on Circular Water-energy-nutrient Systems PrAEctiCe will provide a novel agro-ecology indicator set for East Africa, aimed at helping smallholder farmers in their agro-ecological transition. Financed by EU (Horizon Europe Programme) The project goes beyond the existing indicator frameworks by putting the “concept into action” with a decision support tool for agro-ecology advisors supporting the selection of the best suited Duration combination of agro-ecological practices in a local context. In addition, it puts a focus on circular water-energy-nutrient systems of 2022 to 2026 integrated aqua-agriculture, and practice with high potential for efficient farming with minimal climate impacts, which has not been sufficiently explored in previous indicator work. Through a Partners multi-stakeholder approach, new insight on agro-ecological practices in UM Faculty of Electrical Engineering East Africa will be gathered to inform on existing successful and Computer Science, Hochschule practices as well as the barriers and drivers of East African smallholder Karlsruhe, Steinbeis 2I GmbH, farmers. This insight will help develop an indicator framework for agro-Aquabiotech Limited, Prototipi Limited, ecology, which, while building on existing frameworks, is adapted to the Goeteborgs Universitet, Makerere East African context and captures integrated aquaagriculture practices University, Uganda Martyrs University, in detail. Ministry of Agriculture Livestock and Fisheries Kenia, Regional Universities The PrAEctiCe decision support tool will then, at the farm level, help assess Forum for Capacity Building in environmental and socioeconomic impacts, with a particular focus on Agriculture, National Agricultural impacts on climate change mitigation and adaptation as well as financial Research Organisation, Maseno viability. The tool will be validated in three living labs, situated in Kenya, University, Sustainable Agriculture Uganda and Tanzania, covering different integrated aqua-agriculture Tanzania, Aquagri, Alliance for Food farming set-ups. Knowledge sharing activities through trainings, student Sovereignty in Africa, Africa Agribusiness exchanges and events, ensure the dissemination of results across East Academy Africa and between African Union and EU. To reach practitioners at every level, a cascade training mechanism with a train-the-trainer course will help agro-ecology advisors train farming representatives at the local level who then will help the farmers in their agroecological transition. Policy recommendations for African Union and EU policies will round off the project. 86 GeMMA Activity Report 2016–2022 Green.DAT.AI – Energy-efficient AI-Ready Data Spaces GREEN.DAT.AI aims to channel the potential of AI towards the goals of the European Green Deal, by developing novel Energy-Efficient Large-Scale Data Analytics Services, ready-to-use in industrial AI-based Financed by systems, while reducing the environmental impact of data EU (Horizon Europe Programme) management processes. GREEN.DAT.AI will demonstrate the efficiencies of the new analytics Duration services in four industries (Smart Energy, Smart Agriculture/Agrifood, Smart Mobility, Smart Banking) and six different application scenarios, 2022 to 2025 leveraging the use of European Data Spaces. Partners The ambition is to exploit mature (TRL5 or higher) solutions already developed in recent H2020 projects and deliver an efficient, massively UM Faculty of Electrical Engineering distributed, open-source, green, AI/FL - ready platform, and a validated and Computer Science, Inlecom go-to-market TRL7/8 Toolbox for AI-ready Data Spaces. The services Innovation Astiki Mi Kerdoskopiki will cover AI-enabled data enrichment, Incentive mechanisms for Data Etaireia, University of Piraeus Research Sharing, Synthetic Data Generation, Large-scale learning at the Edge/Fog, Centre, Consiglio Nazionale Delle Federated & Auto ML at the edge/fog, Explainable AI/Feature Learning Richerche, Konnecta Systems Limited, with Privacy Preservation, Federated Atos IT Solutions and Services Iberia & Automatic Transfer Learning, Adaptive FL for Digital Twin Applications, s.l., Erevnitiko Panepistimiako Institouto Automated IoT event-based change detection/forecasting. Tilepikononiakon Systimaton, ITC - Inovacijsko tehnološki grozd Murska The GREEN.DAT.AI Consortium consists of a multidisciplinary group of Sobota, Caixabank SA, Ferrovial 17 partners from 10 different countries (and one associated Servicios SA, Aegis IT Research party), well balanced in terms of expertise. The vast majority of partners GmbH, Red Hat Israel Ltd., Inesc tec - already have key roles in a number of projects funded under Instituto de Engenhariade Sistemas e the Big Data PPP (ICT-16-2017) topic, namely BigDataStack, CLASS, Computadores, Tecnologia e Ciencia, Track & Know, and I-BiDaaS and are serving as active members of Waboost razvoj tehnologij d.o.o., CNET the BDVA/DAIRO Association, FIWARE, AIOTI, and ETSI. In addition, Centre for New Energy Technologies partners come from a variety of sectors, such as banking, mobility, SA, Intasoft International SA, SUNESIS, energy, and agriculture, constituting a representative workforce of their inovativne tehnologije in storitve,d.o.o., respective domains, which will contribute to industry Sphynx Technology Solutions AG adoption and stimulate uptake in other sectors as well. 87 GeMMA Activity Report 2016–2022 Figure 40: GREEN.DAT.AI vision & pathways towards impact at a glance.(Source: own) 88 GeMMA Activity Report 2016–2022 INDY – Energy Independent and Efficient Deployable Military Camps This is a European Defence Fund project lead by Laboratory for power engineering at UM FERI, where GEMMA lab. is a collaborating partner. Both at national and international levels, the energy transition is high Financed by on the priority list. In line with the “Green Deal”, the greening of military EU (European Defence Fund) forces is necessary to reach the ambition defined by Member States. At the moment, deployable military camps are almost 100% dependent on fossil fuel. This not only means that the environmental footprint is high Duration but it also represents a weak point and vulnerability for the armed forces in terms of logistics and dependence on fossil fuels. This is becoming 2023 to 2025 even more challenging given the growing need of energy for the military equipment. INDY proposes a roadmap toward energy independent and efficient Partners systems for military camps. UM Faculty of Electrical Engineering and Computer Science, TECES, The necessary energy transition for deployable military camps will Tehnološki center za električne stroje, require a change of paradigm for energy production, conversion, AVL List GmbH, CAFA TECH OU, CNV storage, transport, distribution and usage, with the final goal of a total CONSULTING, Commissariat à l’Energie independency of fossil fuels. The ambition of INDY is to define a new Atomique et aux Energies Alternatives, approach to energy as a whole from energy production to its final usage. Equipos Móviles de Campaña ARPA, SAU, Fraunhofer Gesellschaft, INDRA INDY’s expected outcome is a strategic roadmap, based on technological SISTEMAS, SOCIEDAD ANÓNIMA, INEO and methodological studies, for the development and implementation of DEFENSE, INSTITUTO NACIONAL DE disruptive and new energy sources, the management of resources and TECNICA AEROESPACIAL ESTEBAN optimization of needs of military deployable camps. It will be a first step TERRADAS, Institutt for energiteknologi, to a better security of supply of energy and the transition to renewable INTRACOM DEFENCE SINGLE MEMBER energy sources for future military camps. S.A., JOHN COCKERILL SA, KOLEKTOR sETup , storitve energetskega upravljanja , d.o.o, Leonardo Società per azioni 89 GeMMA Activity Report 2016–2022 International bilateral projects 90 GeMMA Activity Report 2016–2022 Research on Abnormal Behaviour Detection and Warning in Real-time Video Surveillance Based on Multimedia Algorithms With increasing importance of social public security and rapid development of video surveillance, social public security depends more and more on video image surveillance systems. Video image surveillance Financed by system has become important indispensable infrastructure for public ARRS – Slovenian Research Agency security. Traditional video surveillance systems only display and record (contract BI–CN/14–15–007) video information. Methods of data processing are usually done by real-time manual monitoring or manual postprocessing. Real-time Duration analysis cannot be done, especially during the unexpected or abnormal events. Moreover, data amount accumulates to terabytes of recorded 2015 to 2016 videos. Because of large amounts of video cameras, it is difficult to find abnormal events or accidents in a reasonable time, whilst querying useful information from the video database. Partners UM Faculty of Electrical Engineering In order to eliminate any potential accident danger and handle the events and Computer Science, Dalian Minzu without any delay, we have collaborated on solving this issues through the University, China bilateral project. We extracted key data from video surveillance system and analysed important information. We have put emphasis on state-of-the-art technology and algorithms for real-time video surveillance that take the advantages of image processing and machine learning. Our goal was to automatically analyse real-time video information input, extract the foreground images and update background images, detect and track moving targets, analyse the tracking targets, and perform an early warning in case of abnormal behaviour. This is the initiative and intelligence of video surveillance system. These algorithms provide 24 hours real-time monitoring and intelligent analysis of the captured information. In case of abnormal cases, the system provides an alarm in time to avoid possible accidents. Hence, such system shall save material and financial resources required of employing monitor workers. 91 GeMMA Activity Report 2016–2022 Pruning by Numbers: Integrating Tree Physiology with Growth Simulation and 3D Reconstruction to Optimize Apple Tree Pruning Tree pruning is one of the most important measures to ensure high yield performance by maintaining a balance between vegetative and reproductive growth. This task is at each year before the growing season Financed by carried out manually by tree growers and requires a deep understanding ARRS – Slovenian Research Agency of tree physiology to predict a tree response to the pruning and years of (contract BI–US/17–18–012) practice. To increase the understanding of tree reactions to pruning, the appropriate computer models were developed very early. The first such model can be found in 1996 when the virtual reality application for apple Duration tree growth after the pruning was presented. 2017 to 2018 Since then, a lot of simulators have been developed for various types of fruit trees (e.g. peach, cherry, and walnut), but apple trees have by far retained the most attention of researchers. Recent software simulators Partners took full advantage of graphical processing units, which enables UM Faculty of Electrical Engineering realistic real-time tree visualization. A good example of new generation and Computer Science, Purdue simulators is IMapple, which incorporates a precise functional-structural University , USA tree growth model for Golden Delicious apple trees, based on long-time measurements with a photorealistic tree visualization employing 3D manifold watertight meshes for representing the tree geometry, together enabling interactive tree growth simulation including flowering and fruiting. Interactive simulation and faithful 3D geometric models is also offered by EduAPPLE, an interactive teaching tool for apple tree crown formation, where user trains a one-year-old apple tree in the form of an unbranched whip towards the desired tree form using different tree training techniques. The growth model in EduAPPLE incorporates only the most basic tree growing rules, common to all cultivars while growing conditions are modeled with random variables to maximize tool generality. However, in order to improve users' pruning skills, this is not enough. To do that, the teaching tool has to take a more active role and start to suggest to the user which branches have to be removed in order to achieve the best results. The first step in that direction represents the work, where the pruning is presented as a combinatorial optimization problem of performing the cuts on a virtual tree model in order to achieve the best light distribution inside the tree crown, which is a good start, but not sufficient to actually be used as a recommendation system. Recent advances in 3D geometric modeling have shown an unprecedented precision in detailed reconstruction, inverse modeling, and physics-based responses of vegetation. The advancement of modern GPU and recent algorithms for vegetation modeling allowed realistic and interactive modeling and simulation of plants at scales and geometric details that were not possible before. 92 GeMMA Activity Report 2016–2022 The purpose of this bilateral project is to combine the approaches in order to develop new pruning optimization software that would not only be used as a teaching tool but also a pruning recommendation system for seasoned fruit growers who want to improve their pruning techniques. To achieve that, we have to integrate tree physiology knowledge collected over long-time observations into EduAPPLE and develop a pruning recommendation system based on criteria resulting from this new knowledge. The expected results would provide a novel approach in the training of apple tree pruning to all those who would like to acquire pruning knowledge in order to ensure local fruit source from their gardens, as well as the virtual assistant to all those who want to optimize apple tree pruning in their orchards. In the scope of the collaboration, we developed a new two-step algorithm for dormant apple tree pruning. In the first step, the tree is shaped into one of the predefined primary forms, e.g. cone or cylinder. After that, the Discrete Differential Evolution is used to additionally remove the branches and optimize the tree light intake in the process. The algorithm has been tested on virtual trees inside EduAPPLE simulator. Figure 41: Two-step virtual pruning process giving the control over tree height and neighboring distance. (Source: own) 93 GeMMA Activity Report 2016–2022 Unlike related algorithms based on Differential evolution, our algorithm is capable of preserving distance between neighboring trees in the orchard, and it can control the tree height as well. The simulation inside EduAPPLE showed that the developed algorithm is capable of autonomous tree training towards desirable growing form by pruning the trees over a period of time. Figure 42: Tree training of five apple trees into a Slender Spindle growing form for six consecutive years. (Source: own) 94 GeMMA Activity Report 2016–2022 Research on Intelligent Early Warning of Emergency from Perspective of Public Security Outline of national development plan for science and technology in medium- and long-term points out that public security is the cornerstone of national security and social stability. We are facing big public security Financed by threat in China, which proposes important strategic requirements for ARRS – Slovenian Research Agency (conscience and technology. To cope with this challenge, our government tract BI–CN/18–20–0001) needs to construct one technology system of public security to prevent or control the emergency and adopt efficient response effectively. To bring Duration about a multi-functions integrated emergency guarantee pattern with information and intelligence technology application as the leader, this 2018 to 2020 system will help our government to strengthen its capability of dealing emergencies with rapid response and by predicting sudden events scientifically. From the perspective of public safety, this project utilizes Partners video data captured by cameras installed in public areas and does early UM Faculty of Electrical Engineering warning for crowd abnormal activities, as well as providing service for and Computer Science, Dalian Minzu public safety of ethnic minority areas. University, China Although there are many important research theories and applications in video behavior recognition in recent years, but there are only a limited number of research works related abnormal behavior based emergency from perspective of public security, especially for the performance of real-time or robustness of abnormal crowd behaviors recognition algorithms. There is still a lot of research work to do in early warning for sudden group abnormal behavior. In the figure below a prototype system example developed within the project is shown, where pedestrians are detected by using convolutional neural networks (CNN) deep learning approach. Figure 43: Developed approach for pedestrian detection using CNN (Source: S. Liu et al., Pedestrian Detection based on Faster R-CNN International Journal of Performability Engineering, 15 (7), pp. 1792, 2019). 95 GeMMA Activity Report 2016–2022 Hologram Representation of Cultural Heritage Curators of museums, archaeological sites, architectural and other cultural monuments became conscious, decades ago already, of Financed by capabilities of digital technologies to improve perception of cultural heritage (CH) among visitors of such sites and to enable the perception ARRS – Slovenian Research Agency to interested remote audience. Many artefacts are not permanently (contract BI–BA/1920–003) available to visitors due to the lack of exhibition space, restoration work, or the artefacts’ fragility. Digital technologies may also reassemble broken or otherwise damaged CH monuments, supplement them with Duration the missing (virtual) parts, and embed them in a wider historical context. 2019 to 2020 They also enable fast and safe exchange of contents among distinct institutions, both in a form of visiting virtual exhibitions or for completion of their own material collections with reasonably related digital contents. Partners The digitalisation and visualisation of CH thus represent important stimulators for further progress of spatial data capture and modelling, UM Faculty of Electrical Engineering computer graphics, virtual and augmented reality, human-computer and Computer Science, University of interaction and digital multimedia. Through the previous cooperation, Sarajevo, Bosnia and Herzegovina both groups focussed on geometric modelling, interaction with virtual environments and spatial data input. This time, the partners aimed to address the visualisation of the digitalised CH. The visitors of CH sites are usually not satisfied with traditional visualisation technologies. They expect richer 3D experience through alternative output devices and new ways of interaction. In this project, the challenges of utilisation of holographic pyramids in applications of CH visualisation were addressed. Though these devices do not provide real holograms but only a hologram-like visual illusion, the technology gained huge popularity. Miniaturised implementation, usually combined with a mobile device, is easily accessible to everyone, while bigger professional installations may elegantly, attractively and functionally supplement the museum interior due to their futuristic shape and particularly the created 3D illusion. UM FERI has detected some open challenges related to the preparation of digitalised contents, design of the device as a whole, its hardware and optical components in order to increase the realism of created illusions. ETF Sarajevo (Faculty of Electrical Engineering, University of Sarajevo) utilized its knowledge and practical experience with the CH visualisation, performed studies of user requirements and analyses of user experience in order to set up the directions for validation of the considered technological solutions. We also assessed particular experimental installations regarding the set directions. Particular challenges within preparation of data were projected in the holographic pyramid comprising inclusion of bigger scenes, background, terrain and avatars (3D human models which guide a visitor through the virtual environment, talk stories about the artefacts etc.). Namely the contents to be displayed in holographic pyramids are usually prepared in a manner to produce an illusion of a single object floating within the pyramid. There are eventually several smaller objects displayed, but they are still positioned in a manner providing the projection to the middle of the pyramid. However, a wider context is usually desired when the CH is being considered. 96 GeMMA Activity Report 2016–2022 Figure 44: Interactive control of the digitized Roman artefact, found in Balkan area in holographic pyramid with the use of Leap motion sensor. (Source: own) Thus, an object should placed in a concrete space, onto the terrain, next to other objects and in front of the background representing its original historical location. We also prefer an avatar somewhere in the peripheral part of the pyramid as it is aimed to supplement the exhibited artefact and not to obstruct or substitute it. A part of the UM FERI group deals with the construction of holographic pyramids. The listed challenges shall, thus, be also addressed through experimenting with different organisations of optical components and with utilisation of various materials with diverse optical characteristics. The holographic pyramids for the so-called integral photography are often used as an alternative to the regular holographic pyramids. The research group from ETF Sarajevo provided adequate geometric models to be visualised within the pyramids, and also assessed the user experience due to the determined goals and criteria. A user experience obtained through organisation of the scene into a system of multiple holographic pyramids or by combining the pyramids with other elements of immersive environments, narration and scenography may also represent an interesting challenge. The interaction also plays an important role in achieving the overall user experience. To support eventual future complete system implementation, we also studied the manipulation of represented contents with manifold technologies of virtual and augmented reality. 97 GeMMA Activity Report 2016–2022 Figure 45: Digitized Roman artefact, found in Balkan area, visualised, and controlled through Microsoft HoloLens glasses. (Source: own) 98 GeMMA Activity Report 2016–2022 Analysis of Internet-based Cultural Transmission by Knowledge Graphs Data mining algorithms are becoming important for the analysis of big data such as social networks graphs, where cultural transmission is one of the semantic components. One of the exciting new algorithms are Financed by based on knowledge graphs, which were firstly proposed by Google in ARRS – Slovenian Research Agency 2012. Knowledge graph generally refers to an algorithm for constructing (contract BI–CN/20–21–20) a subgraph representing semantic relation network. Currently most of the researches focus on the theoretical research of culture transmission, Duration or on the algorithmic aspects of knowledge graph construction. Few scholars combine the two aspects, in order to study how to realize the 2020 to 2021 sustainable development of culture transmission. Based on the cross media big data of culture transmission, this project fully takes the advantages of big data mining and parallel computing technology. The Partners aim would be based on construction of knowledge graphs of culture UM Faculty of Electrical Engineering transmission and extraction of public interest points. Relying on big and Computer Science, Dalian Minzu data parallel computing technology, we will enhance the performance of University, China cultural transmission analysis. The project will also have social impacts beyond the scope of the project, by disseminating the attraction of internet-based cultural transmission, while demonstrating the influence of Chinese and Slovene cultures algorithms. In the figure on the left, the developed methodology is shown, where knowledge graph is constructed from source texts, and then used in answering generation module. Figure 46: Automatic answering system based on knowledge graph (Source: S. Liu et al., Research on Automatic Question Answering of Generative Knowledge Graph Based on Pointer Network, Information, 12.3, 136, 2021). 99 GeMMA Activity Report 2016–2022 Industrial projects 100 GeMMA Activity Report 2016–2022 TunePerfect Goal of the project was to develop simple UI/UX for self-fitting of the hearing aid devices as an alternative to the audiologist driven fitting approach. This was achieved by grid based user exploration where users Financed by can boost hearing aids' frequency gains (EQ) by navigating though the presented grid which on interaction temporarily applies EQ directly to Altran Switzerland AG the connected hearing aid. Therefore, by exploring the grid matrix, one can adjust properties of the hearing aid device and find optimal hearing for the provided audio samples. User is guided through 4 different steps Duration where first step sets the global EQ curve and other steps serve as fine 2016 to 2017 tuning of the EQ. Our involvement with the project was to develop: Partners • Friendly interactive tutorial which guides users through self-fitting process; UM Faculty of Electrical Engineering and • Tool that enables 3rd parties to translate every aspect of the Computer Science, CwIT s.p. application into their desired language; • Tool that performs analysis based on TunePerfect usage telemetry and visualizes user interaction with the software in form of heatmap Additional information which enables further UX adjustments. http://audioap.ciopro.si/ Figure 47: TunePerfect integrated into PhonakTarget application developed by Sonova. (Source: own) 101 GeMMA Activity Report 2016–2022 Design of the Information Support for Vegetation Management in the Power-line Corridors Vegetation poses a danger to power lines, as trees can fall on them. Therefore, it is needed to perform periodic inspection of vegetation and Financed by perform appropriate actions. The aim of this study was the development of a plan for an automated vegetation management system and its ELES d.o.o. integration into the existing information system for the Slovenian Electricity Transmission System Operator – ELES. Duration The most important functionalities of the system are the following: 2017 to 2018 • Detection of trees; • Transparent management of vegetation data and restrictions for taking actions on the vegetation; • Forecasting the vegetation development; • Support for planning measures and tools for analysis and reporting on vegetation management workflow. On the basis of the survey of remote sensing technologies, possible ways of using them for the implementation of the predicted functionalities of the vegetation management system were proposed. The study proposed a separate vegetation management module that connects with the existing components of the information system through existing databases and common data structures. Finally, the study provided an estimate of the costs and benefits of the potential introduction of the vegetation management system. Figure 48: Components of the information system for vegetation management on transmission line corridors. (Source: own) 102 GeMMA Activity Report 2016–2022 Development of Geoinformation Systems, Services and Solutions The aim of the project was the development of the geoinformation systems and services, which are divided in these three areas: Financed by • Development of algorithms, libraries and visualizations for use in Java application for editing geometric objects, which are stored in Igea d.o.o. the relational database Oracle 12c and accessed through the spatial infrastructure (Geoserver) and dedicated REST services in the form of GeoJSON and TopoJSON. Editing of geometric objects (point, Duration line, polygon, polyline and multipolygon) is performed in the way of 2018 to 2021 minimal changes and optimized geometry construction on server and client side with traceability of changes provided (versions of geometric objects); • Development and upgrades of system tools with INSPIRE development guidelines for the management of metadata structures and descriptions of spatial and descriptive data based on open-source solutions Geonetwork and Re3gistry (shown on image bellow).; • Development of tools for monitoring and managing data flows of heterogeneous sensor systems. Figure 49: Slovenian system of registries and codelists. (Source: own) 103 GeMMA Activity Report 2016–2022 Advanced Visualization Components GEMMA has collaborated with the DEWESoft company and developed advanced solutions for transfer and 2D/3D visualization of geographic Financed by data based on OGC standards. These solutions were integrated into DEWESoft software as custom visual controls which implement the DEWESoft d.o.o. necessary interfaces to be consumed by the application. Developed solution enables users to synchronize geolocated measurements on Duration 2D or 3D map (e.g. real-time tracking of transport vehicles), insertion of custom 3D models, and tracking of different vector or scalar parameters 2018 to 2022 (e.g. color-mapping the visual track based on velocity change). After successful release and positive feedback of map component, demand for aviation features increased, resulting in 3D terrain support and standalone 3D model visual control. New control can be used for aircraft orientation visualization without need for the geographic location. Furthermore, Map component was upgraded to support the real-time LiDAR data visualization from mounted hardware. Figure 50: 3D Map of the Hockenheimring racing track. (Source: own) 104 GeMMA Activity Report 2016–2022 Figure 51: 3D map with terrain details and 3D model visual controls overlayed on top of each other. (Source: own) Figure 52: Real-time LiDAR data visualization without active map layers and selected 3D car model. (Source: own) 105 GeMMA Activity Report 2016–2022 Since then, GeMMA also shifted to the non-geospatial fields of expertise and helped to improve the existing visual control for the modal geometry which is indispensable tool for understanding the vibration aspect of mechanical structures through visualization and animation. Geometry can either be loaded from standard UNV file format or created manually in geometry editor, where vertices can be mapped to the real measurement data in order to animate entire structure. Moreover, the textbox widget was developed, which supports rich text formatting capabilities and allows for custom user annotations anywhere on the screen. Additionally, it supports user expressions for displaying real-time values and properties from the data channels. Figure 53: Modal geometry and text widget in DEWESoft software. (Source: own) Additional widgets to help with the time domain and frequency domain data visualization were also developed to be used together with orbit analysis module that is used for vibration analysis of rotating machinery. Figure 54: Waveform graph and polar plot widgets used with orbit analysis module outputs. (Source: own) Up till 2022, GeMMA lab. was involved in the development of 9 different visual controls, which are part of the core package and is actively contributing to the DEWESoft core software by improving its 3D graphics engine using DirectX graphics library. 106 GeMMA Activity Report 2016–2022 ORYX – Massage Roller Control Application Mobile application for controlling Bluetooth® enabled Relaxroll devices. Connect your Relaxroll devices, start a routine, and let the application do the thinking and just relax. Tap into curated routines and enjoy automated Financed by speed setting control. Relaxroll Gmbh ORYX is a foam massage roller with integrated Bluetooth® chip for communication with other devices. The aim was to develop a frontend and backend application to control smart wellness devices. Duration 2019 The application is able to: • Notify the user for recovery sessions; • Suggest a predefined recovery session; • Connect to smart devices; • Changning speeds of the connected device; • 3rd party connectivity. Figure 55: Oryx application. (Source: own) 107 GeMMA Activity Report 2016–2022 Advanced Investigation Platform The Advanced Investigation Platform (AIP) allows the user to search and analyse patterns in big network data. The entire platform consists of four Financed by core levels, namely: • Presentation; MNZ RS, Police • Application; • Logic; • Data levels. Duration 2019 The presentation level consists of a web application that allows a user to define simple and advanced queries via the user interface, or directly using a script. The web application also supports the visualisation of query results, and highlights a set of functionalities from the application level in a user-friendly way. It implements an Application Programming Interface (API) between the user and the logical layer in the form of REST service queries. The logical level of the platform consists of data level access, security, authorisation, authentication, algorithms for statistical analysis and finding correlations in data samples, and support for adding various external services that enable scalability of the platform (e.g. mapping between arbitrary data formats). The data layer consists of various encrypted databases, such as a user database, a database of networks, a database of registered plug-ins, a database for replicating and buffering query results, and a database for storing analytical results. 108 GeMMA Activity Report 2016–2022 Development of an Application for Interactive Presentation of Museum Exhibitions Within the laboratory, we developed an application for the interactive presentation of the exhibits of the museum collection. The application serves as the innovative replacement for traditional info panels usually Financed by accompanying the exhibitions. Instead of a classical static presentation, the application enables the presentation to be broken into a series Maribor Regional Museum of stories bound to the exhibit, which the visitor can browse through. Each story is centered around the photograph, or a picture displayed in a special bar next to the exhibit's description. The entire collection Duration is represented with images displayed in one or more image stripes. 2020 The corresponding database and module for data entry support the application. The application is a part of the permanent exhibition titled Spaces of the beautiful in Maribor Regional Museum. Figure 56: Photograph and description of the exhibition object, together with related objects within the Application for Interactive Presentation of Exhibits in the Museum Collection. (Source: own) 109 GeMMA Activity Report 2016–2022 Proposal for the Development Plan of the Spatial Data Infrastructure of the Municipality of Maribor The purpose of the MOM (Municipality of Maribor) Spatial Data Infrastructure Development Plan is: Financed by • Guidelines for individual projects and investments in spatial data infrastructure for the needs of municipal authorities, funds, urban Municipality of Maribor districts and local communities, as well as public utilities and public institutes; Duration • Coordinated implementation of projects for the establishment of spatial data and the development of information equipment; 2020 to 2021 • Organizational chart in which the actors, their roles and duties will be identified, which will be the basis for the organization of the field. The following activities were carried out: • Preparation of materials and conducting interviews; • Analysis of the situation, needs and development plans by departments of the city administration and public service providers and other users of spatial data and services; • Analysis of trends and needs arising from the SUS (Sustainable Urban Strategy), the Smart City of Maribor initiative, the national project e-space and legislation (emphasis on ZUreP-2, GZ, Environmental Protection Act) and at least one case from abroad; • Proposal of the plan for the development of the spatial data infrastructure of the Municipality of Maribor; • Preparation of materials and implementation of workshops with representatives of the Municipality of Maribor. 110 GeMMA Activity Report 2016–2022 Research and Development of Geoinformation Systems, Services and Support Algorithms The goal of this project was the development of information systems, services and solutions using the EMRIS (Unique Methodology of Development of Information Systems) methodologies and agile methods. Financed by This Includes: Igea d.o.o. • Development of java libraries and specific REST services for processing and visualization of GeoJSON and TopoJSON data, as well as optimized storage in the Oracle 12c database using the Duration GeoServer spatial infrastructure; 2021 to 2022 • Development of algorithms for geometrical object editing based on »minimal change« paradigm; • Development of optimized algorithms for tracking data change on the server and on client side; • Development of tools for managing metadata using the INSPIRE guidelines based on opensource tools GeoNetwork and Re3gistry; • Development of libraries for spatial raster data processing; • Development of tools for processing of data streams from sensory systems. Figure 57: Slovenian INSPIRE Data Portal. (Source: own) 111 GeMMA Activity Report 2016–2022 VegeLine – Information System for Risk Assessment and Vegetation Management inside Powerline Corridors Project goal was to develop and deploy information system for risk assessment and vegetation management which aims Financed by to reduce manual terrain surveying and increase long-term planning efficiency for vegetation clearance inside powerline corridors. Main ELES d.o.o. objective was to detect critical areas and prevent risk of the potential network outages caused by excessive vegetation growth, with secondary Duration ideal to optimize for lower operational costs of the internal and external vegetation management services. Overall, the goal was to improve 2021 to 2022 planning efficiency of the vegetation clearance schedules over longer time periods. Partners The main functionalities of the developed information system are calculation of the vegetation growth based on predictive canopy UM Faculty of Electrical Engineering and growth models, detected risk areas, and algorithms for intervention Computer Science, Troia d.o.o., Inova IT optimizations based on estimated operational costs. d.o.o. Figure 58: Visualization of the spatial layers produced by the vegetation management system. (Source: own) 112 GeMMA Activity Report 2016–2022 To yield mentioned results, the system incorporates wide range of available spatial data from integrated first party services in form of temporal LiDAR scans, powerline networks, protected areas for species and habitats preservations, and environmental data such as amount of precipitation, air temperature, sunshine duration, and soil qualities. System then automatically produces and derives necessary products such as digital terrain and canopy height models, spatial filters which define maximum vegetation heights, and outputs of the vegetation growth simulations. Derived data is required for further processing and decision making and results of the system are detailed risk assessments and clearance plans which aim to minimize operational costs. System exposes those results in form of standardized services and formats which are then consumed from within existing IBM Maximo asset management system and ESRI ArcGIS. Moreover, advanced client tools were developed for system administration, data-source management, raw data editing, visual data analytics and detailed intervention planning capabilities. Figure 59: Risk assessment and tree growth predictions for 10 years into the future on the selected powerline span. (Source: own) 113 GeMMA Activity Report 2016–2022 Assessment of the Photovoltaic Potential of Pošta Slovenije's Real Estate Facilities This is an R&D project financed by Pošta Slovenije d.o.o., where Laboratory for power engineering at UM FERI jointly collaborated Financed by with the GEMMA lab. The project provided an insight into photovoltaic capacity of Pošta Slovenije's real estate facilities. Due to a high number Pošta Slovenije d.o.o. of facilities the financer required a systematic overview of photovoltaic potential assessment to support decisions regarding the investment into photovoltaic systems. An automatic system for generating reports Duration regarding the assessment of photovoltaic potential of any building in Slovenia was developed. The system automatically obtains LiDAR 2021 data from a public database and processes it for each location to take shadowing from surroundings into account. The measurements of direct and diffuse radiation from the closest meteorological stations of each location were used for the calculation of solar radiation. The reports include general information of each real estate, mapped influences of each considered assessment factors (aspect, inclination, shadowing, …), solar potential and photovoltaic potential. For the assessment of photovoltaic potential the characteristics of three solar panels that were provided by the financer were considered. In addition to the dimensions of solar panels, the non-linear characteristics that affect efficiency in relation to the received radiation were considered at an hourly time step. The electricity production was provided on a monthly level for each type of panel as shown in the figure below. Figure 60: Monthly electricity production for a selected real estate facility. (Source: own) 114 GeMMA Activity Report 2016–2022 The photovoltaic potential was first given for a case when the whole roof would be used for solar panel installation as a measure for total capacity. The maximum capacity was given as rated power and estimated using the number of possible installations of solar panels in regard to the dimensions of each panel. Additional case for improved return of investment was provided by excluding the highly shaded roof areas, as shown in the following figure. Figure 61: A selected real estate's received solar radiation for a case with excluded highly shaded roof. (Source: own) 115 GeMMA Activity Report 2016–2022 Architecture and Interoperability Management Plan for Spatial Information The Ministry of the Environment and Spatial Planning (MOP) is the holder of a multitude of spatial data within the state infrastructure, as it manages over one hundred databases in the field of space, environment, Financed by nature, water and real estate, which serve as a source of reference for MOP RS, The Surveying and mapping other users, however, it is not the only manager and provider of spatial Authority of the Republic of Slovenia data. Within the Ministry, there are different areas of work within the directorates, offices, agencies and constituent bodies. As a result, various internal and external data flows are emerging, both internally Duration and nationally, as well as in interaction with the EU (eg environmental 2021 to 2022 reporting, INSPIRE-2 transfer). From the point of view of information architecture, organization of informatics and spatial data infrastructure, there is no single, validated model that would provide effective support within the ministry for future information society requirements and upcoming projects within the ministry and wider at the national level. This project provided guidelines and a plan for the establishment of an internal infrastructure for the management of spatial data of the Ministry, as well as guidelines and a plan for the establishment of a common national spatial infrastructure. In other words, the project provided a blueprint for the implementation of dataspace in the geospatial domain. The technical guidelines and the establishment plan was based on existing strategic and implementation documents created in the field of spatial information management, which seek to identify gaps that arise in the implementation of integration in the future. The purpose of the document was to get acquainted with some trends and variants of solutions and concepts that ensure the stable operation of the infrastructure in the future in accordance with the areas of application. The project also resulted in the analysis of options for providing information infrastructure and the development of a plan for the establishment of a common spatial information infrastructure of the Ministry of the Environment in the field of space, environment, nature, water and real estate. As a result we proposed an organizational and information architecture that the Republic of Slovenia could use in developing and strengthening its approaches to national spatial information management. A plan for the establishment of a common spatial information infrastructure in Slovenia was prepared, with special emphasis on the processes and activities within the competence of the Ministry of the Environment and Spatial Planning. The project work resulted in substantive, organizational and technical guidelines for the development and establishment of the Spatial Information Infrastructure (SII) of the MOP and the Common National Spatial Data Infrastructure at the state level. 116 GeMMA Activity Report 2016–2022 Short-term and Long-term Traffic Forecasting System Using Artificial Intelligence Algorithms The purpose of the project is to create a system that will enable advanced calculation and insight into traffic forecast in accordance with the operational requirements of the client, based on advanced methods of Financed by artificial intelligence. In this context, the development of cloud analytical DARS d.d. algorithms is planned, together with appropriate user applications and their integration into the operational environment of the control center. The project covers the entire motorway network in Slovenia and the Duration most important parallel and main roads. The system will enable short-term, medium-term and long-term forecasting and will be designed to 2021 to 2023 relieve the traffic, to increase the level of autonomy in the analysis, to supplement and improve the quality of possible automated calculations and to increase the speed of reliable forecasts. Partners UM Faculty of Electrical Engineering and The following key objectives of the project can be identified as: Computer Science, CreaPro d.o.o. • Objective 1: Establish an infrastructure for the operation of traffic forecasting system applications; • Objective 2: Establish a data model for recording a calendar of holidays in neighboring countries; • Objective 3: Establishment of a data model and application for the collection and maintenance of data from open online sources on weather and other parameters that affect traffic; • Objective 4: Establish a data collection and merging system to support the operation of the forecasting model; • Objective 5: Integration of traffic meters with real-time data; • Objective 6: Data cleansing; • Objective 7: Implementation of a forecasting model for short-term and long-term traffic forecasts; • Objective 8: Data export services and control application showing forecast results; • Objective 9: Testing and implementation of the solution in production. Within the project, GeMMa's main contributions are: • Examination of web APIs through which relevant data for learning a prediction model is accessible; • Preparation of server data models for the transformation of data into a format that will be suitable for their integration into the development tool and data viewer; • Establishment of appropriate web API interfaces and data viewer for access and integration of data relevant to the forecasting model; • Providing a communication model between databases and the development environment; • Analysis of the current state of the art in the field of forecasting models and, more specifically, traffic flow forecasting models; • Learning and testing predictive models. 117 GeMMA Activity Report 2016–2022 Figure 62: Integration of data in GIS application. (Source: own) Figure 63: Real-time predictions shown on map of Slovenia highway with a gradient ranging from green (low traffic density) to red (congestion). (Source: own) 118 GeMMA Activity Report 2016–2022 Figure 64: Counter data statistics through the day in boxplot where red color represents the vehicle count while yellow colors show vehicle velocity distributions for multiple counters. (Source: own) 119 GeMMA Activity Report 2016–2022 Development of Additional Performances of IMINT BLS – BLS ALGO The aim of the project is to develop additional software features for the Belin and Galeb drone systems. This includes the analysis of video Financed by image material obtained by UAV, including detection and recognition of military facilities, detection of persons in the operational environment MORS RS and detection of the movement of various objects in the field in near real time. Duration Existing BLS capacity (Belin and Galeb) in conjunction with IMINT 2021 to 2022 capacity are an integral part of ISTAR equipment and are intended for obtaining and processing intelligence from the operational environment. The existing IMINT software provides image processing mainly at the Partners operational and strategic level, but does not contain functionalities that are important for the commander at the tactical level: UM Faculty of Electrical Engineering and • Detection of objects or targets; Computer Science, Igea d.o.o., C-Astral • Focusing on military facilities and targets; d.o.o., MIL Sistemika d.o.o., Onedrone • Real-time motion detection. d.o.o., Skylabs d.o.o., Timtec d.o.o. With the development of this function, we will increase the efficiency of image processing and, consequently, the efficiency of intelligence support. Figure 65: UAV monitoring the field with multiple detected objects. (Source: own) 120 GeMMA Activity Report 2016–2022 Pilot System for Determining Waiting Times at Border Crossings On the border between Slovenia and Croatia, especially during the tourist season, there are longer traffic jams. The main cause is administrative border control, which, even in the case of very fast and efficient work of Financed by the border authorities, with a large influx of vehicles, inevitably causes traffic jams. Several systems (FCD, traffic counters, bluetooth) have been MZI RS tested in the past to estimate waiting times. When reviewing solutions for similar problems around the world, the technology of object recognition Duration through video camera recordings was encountered. Therefore, the aim of this pilot project is to test such a system, which will be the basis 2022 for further planning of improving traffic information on the situation at border crossings. The system supports transmition of genrated data on the actual waiting time using standard protocols via dedicated web interfaces in XML or JSON formats, which allows for the subsequent use in the various client's systems. The developed system does not collect any personal data, license plate data or other vehicle identification data. The current waiting time before crossing the border is estimated every minute. Waiting time estimation is done by detecting an individual vehicle and measuring the actual time that the vehicle takes over a certain distance. Since individual vehicles are detected and tracked on the road, the developed system can detect the vehicles that do not provide relevant travel times, such as stationary vehicles or vehicles that are eliminated or excluded from traffic on the section in question. These vehicles are then filtered out before the final waiting time estimation. The proportion of vehicles in the total traffic that the system can detect is high enough to make the data on waiting times reliable and consequently allows for error detection and correction in all weather conditions (rain, sun, night). Figure 66: Vehicle detection and speed monitoring in action. (Source: own) 121 GeMMA Activity Report 2016–2022 An Upgrade of a Viewer for Slovenian Forest Service The forest data viewer was primarily developed to allow the employees of the Slovenia Forest Service to publish the forestry related data they collect and produce. This data is then made available to the general Financed by public through interactive GIS web application and related OGC compliant Slovenia Forest Service d.o.o. services. The developed solution consists of several modules: Duration 2022 • Backend relational database for storing vector and meta data; • Backend storage for large raster data; • Backend services for OGC compliant data distribution and styled map generation (using open-source software GeoServer); • Backend services for GIS map caching (using open-source software GeoWebCache); • Backend web application specific services (feature identification, search queries, area data export…); • Interactive frontend web GIS application. Additionally, the forest data viewer also provides the services for computing the land plot wood stock information (types of trees, amount of wood and other related information). The user can download the forest management plans (provided by Slovenia Forest Service). Several search options allow the user to quickly locate the area of interest and then export the content as PDF or as an image. Figure 67: Forest data viewer. (Source: own) 122 GeMMA Activity Report 2016–2022 ST Publications Prizes, Awards, Honours, Medals Ph.D. Candidates Granted by ARRS and Completed Ph.D.s Supervised in GeMMA CHIEVEMEN Functions and Honours in National and International Associations A 123 GeMMA Activity Report 2016–2022 Publications Original Scientific Articles 1. Jesenko, D., Š. Kohek, B. Žalik, M. Brumen, D. Kavran, N. Lukač, A. Živec, A. Pur. STALITA: innovative platform for bank transactions analysis. Applied sciences. 12 (23) 2022, 13. 2. Liu, S., M. Xu, Y. Qin, N. Lukač. Knowledge graph alignment network with node-level strong fusion. Applied sciences. 12 (19) 2022, 16. 3. Uremovič, N., M. Bizjak, P. Sukič, G. Štumberger, B. Žalik, N. Lukač. A new framework for multivariate time series forecasting in energy management system. IEEE transactions on smart grid. 2022, 14. 4. Kohek, Š., B. Žalik, D. Strnad, S. Kolmanič, N. Lukač. Simulation-driven 3D forest growth forecasting based on airborne topographic LiDAR data and shading. International journal of applied earth observation and geoinformation : the journal is the successor of the former ITC Journal and has been published by Elsevier since 2002. 11 2022, 13. 5. Jesenko, D., D. Mongus, U. Lešnik. The Influence of COVID-19 on particulate matter concentrations in a medium-sized town. Promet. 34 (5) 2022, 813-823. 6. Jesenko, D., L. Hruda, I. Kolingerová, B. Žalik, D. Podgorelec. Symmetry-based method for water level prediction using sentinel 2 data. Sensors & transducers. 256 (2) 2022, 12-18. 7. Cukjati, J., D. Mongus, K. Rizman Žalik, B. Žalik. IoT and satellite sensor data integration for assessment of environmental variables: a case study on NO2. Sensors. 22 (15) 2022, 16. 8. Borovič, M., M. Ojsteršek, D. Strnad. A hybrid approach to recommending universal decimal classification codes for cataloguing in slovenian digital libraries. IEEE access. 10 2022, 85595-85605. 9. Žalik, B., D. Strnad, Š. Kohek, I. Kolingerová,A. Nerat, N. Lukač, D. Podgorelec. A hierarchical universal algorithm for geometric objects reflection symmetry detection. Symmetry. 14 (5) 2022, 1-21. 10. Kohek, Š., N. Lukač, D. Strnad, I. Kolingerová, B. Žalik. Data on annotated approximate bilaterally symmetric leaf-off trees based on particle flow simulation and predefined tree crown shape. Data in brief. 40 2022, 1-5. 11. Strnad, D., Š. Kohek. Constrained multi-objective optimization of simulated tree pruning with heterogeneous criteria. Applied sciences. 11 (22) 2021, 1-18. 12. Kolmanič, S., D. Strnad, Š. Kohek, B. Benes, P. Hirst, B. Žalik. An algorithm for automatic dormant tree pruning. Applied soft computing. 99 2021, 1-11. 13. Bogataj, M., Z. Kravanja, A. Soršak, B. Slemnik, U. Klemenčič, S. Jurič. MIPSYN-global: process synthesis enabled by graphical modelling. Chemical engineering transactions. 88 2021, 631-636. 124 GeMMA Activity Report 2016–2022 14. Bizjak, M., B. Žalik, G. Štumberger, N. Lukač. Large-scale estimation of buildings' thermal load using LiDAR data. Energy and buildings. 231 2021, 1-16. Vlahek, D., D. Mongus. An efficient iterative approach to explainable feature learning. IEEE transactions on neural networks and learning systems. 2021, 1-13. 15. Žalik, B., D. Mongus, K. Rizman Žalik, D. Podgorelec, N. Lukač. Lossless chain code compression with an improved Binary Adaptive Sequential Coding of zero-runs. Journal of visual communication and image representation. 2021, 1-17. 16. Bizjak, M., B. Žalik, N. Lukač. Parameter-free half-spaces based 3D building reconstruction using ground and segmented building points from airborne lidar data with 2D outlines. Remote sensing. 13 (21) 2021, 1-17. 17. Mongus, D., M. Brumen, D. Žlaus, Š. Kohek, R. Tomažič, U. Kerin, S. Kolmanič. A complete environmental intelligence system for LiDAR-based vegetation management in power-line corridors. Remote sensing. 13 (24) 2021, 1-15. 18. Tomažič, L. M., N. Lukač, G. Štumberger. . A new regulatory approach for PV-based self-supply, validated by a techno-economic assessment : a case study for Slovenia. Sustainability. 13 (3) 2021, 1-14. 19. Lukač, N., D. Špelič, G. Štumberger, B. Žalik. Optimisation for large-scale photovoltaic arrays' placement based on Light Detection And Ranging data. Applied energy. 263 2020, 1-11. 20. Strnad, D., Š. Kohek, B. Benes, S. Kolmanič, B. Žalik. A framework for multi-objective optimization of virtual tree pruning based on growth simulation. Expert systems with applications. 162 2020, 1-10. 21. Triglav Čekada, M., D. Radovan, B. Lipuš, D. Mongus. Very small glaciers as geoheritage : combining a spatio-temporal visualisation of their development and related effects of climate change. Geoheritage. 12 (85) 2020, 1-11. 22. Žlaus, D., D. Mongus. Efficient method for parallel computation of geodesic transformation on CPU. IEEE transactions on parallel and distributed systems. 31 (4) 2020, 935-947. 23. Žalik, B., D. Mongus, N. Lukač, K. Rizman Žalik. Can Burrows-Wheeler Transform be replaced in chain code compression? Information sciences. 2020, 1-18. 24. Kolmanič, S., M. Marksel, D. Mongus, B. Žalik. Tehnologije navidezne in obogatene resničnosti, kot orodje za predstavitev novih idej in produktov na sejmih : primer Mahepa. Uporabna informatika. 28 (2) 2020, 85-93. 25. Mongus, D., M. Brumen, B. Kozan. Merjenje nadmorske višine gladine jezer iz optičnih satelitskih slik. Uporabna informatika. 28 (3) 2020, 131-138. 26. Liu, S., H. Yang, J. Li, S. Kolmanič. Preliminary study on the knowledge graph construction of chinese ancient history and culture. Information. 11 (4) 2020, 1-21. 27. Žalik, B., K. Rizman Žalik, E. Zupančič, N. Lukač, M. Žalik, D. Mongus. Chain code compression with modified interpolative coding. Computers & electrical engineering. 77 2019, 27-36. 28. Strnad, D., Š. Kohek, A. Nerat, B. Žalik. . Efficient representation of geometric tree models with level-of-detail using compressed 3D chain code. IEEE transactions on visualization and computer graphics. 2019, 1-13. 29. Nerat, A., D. Strnad, E. Zupančič, B. Žalik. Extended algorithm to construct a quadtree from freeman chain code in four directions. Image analysis & stereology : official journal of the International Society for Stereology. 38 (3) 2019, 227-235. 30. Liu, S., P. Cheng, J. Liu, N. Lukač. Piecewise combination of hyper-sphere support vector machine for multi-class classification problems. International Journal of Performability Engineering. 15 (6) 2019, 1611-1619. 125 GeMMA Activity Report 2016–2022 31. Liu, S., X. Cui, J. Liu, N. Lukač. Pedestrian detection based on faster R-CNN. International Journal of Performability Engineering. 15 (7) 2019, 1792-1801. 32. Kohek, Š., D. Strnad, B. Žalik, S. Kolmanič. Interactive synthesis and visualization of self-organizing trees for large-scale forest succession simulation. Multimedia systems. 25 (3) 2019, 213-227. 33. Jeromel, A., B. Žalik. An efficient lossy cartoon image compression method. Multimedia tools and applications. 2019, 1-19. 34. Šarlah, N., T. Podobnikar, D. Mongus, T. Ambrožič, B. Mušič. Kinematic GPR-TPS model for infrastructure asset identification with high 3D georeference accuracy developed in a real urban test field. Remote sensing. 11 (12) 2019, 1-26. 35. Zupančič, E., B. Žalik. Data trustworthiness evaluation in mobile crowdsensing systems with users' trust dispositions' consideration. Sensors. 19 (6) 2019, 1-23. 36. Lipuš, B., B. Žalik. 3D convex hull-based registration method for point cloud watermark extraction. Sensors. 19 (15) 2019, 1-18. 37. Rizman, Žalik K., B. Žalik. Node attraction-facilitated evolution algorithm for community detection in networks. Soft computing. 23 (15) 2019, 6135-6143. 38. Lešnik, U., D. Mongus, D. Jesenko. Predictive analytics of PM[sup]10 concentration levels using detailed traffic data. Transportation research. Part D, Transport and environment. 67 2019, 131-141. 39. Jesenko, D., T. Jagrič, B. Žalik, D. Mongus, V. Jagrič. Vedenjski model neplačila za portfelj kreditnih kartic s pomočjo strojnega učenja. Bančni vestnik : revija za denarništvo in bančništvo. 67 (4) 2018, 39-45. 40. Bizjak, M., B. Žalik, G. Štumberger, N. Lukač. Estimation and optimisation of buildings' thermal load using LiDAR data. Building and environment. 128 2018, 12-21. 41. Kolmanič, S., Š. Kohek, B. Žalik. From computer edit geometric design education to virtual worlds: advantages and pitfalls of computer-based education. Computer applications in engineering education. 26 (5) 2018, 1614-1625. 42. Mongus, D., U. Vilhar, M. Skudnik, B. Žalik, D. Jesenko. Predictive analytics of tree growth based on complex networks of tree competition. Forest Ecology and Management. 425 2018, 164-176. 43. Žalik, B., D. Mongus, N. Lukač, K. Rizman Žalik. Efficient chain code compression with interpolative coding. Information sciences. 439/440 2018, 39-49. 44. Rizman Žalik, K., B. Žalik. Memetic algorithm using node entropy and partition entropy for community detection in networks. Information sciences. 445/446 2018, 38-49. 45. Mongus, D., B. Žalik. Segmentation schema for enhancing land cover identification : a case study using Sentinel 2 data. ITC journal. 66 2018, 56-68. 46. Lipuš, B., B. Žalik. Robust watermarking of airborne LiDAR data. Multimedia tools and applications. 77 (21) 2018, 29077-29097. 47. Rizman Žalik, K., B. Žalik. A framework for detecting communities of unbalanced sizes in networks. Physica. A, Statistical mechanics and its applications. 490 2018, 24-37. 48. Kohek, Š., D. Strnad. Interactive large-scale procedural forest construction and visualization based on particle flow simulation. Computer graphics forum. 37 (1) 2018, 389-402. 49. Strnad, D., Š. Kohek, S. Kolmanič. Fuzzy modelling of growth potential in forest development simulation. Ecological informatics. 48 2018, 80-88. 126 GeMMA Activity Report 2016–2022 50. Jesenko, D., M. Mernik, B. Žalik, D. Mongus. Two-level evolutionary algorithm for discovering relations between nodes' features in a complex network. Applied soft computing. 56 2017, 82-93. 51. Rebolj, D., Z. Pučko, N. Čuš Babič, M. Bizjak, D. Mongus. Point cloud quality requirements for Scan-vs-BIM based automated construction progress monitoring. Automation in construction. 84 2017, 323-334. 52. Rizman Žalik, K. Community detection in networks using new update rules for label propagation. Computing. 99 (7) 2017, 679-700. 53. Kolednik, D., D. Mongus, B. Žalik. Postopek zaznave sprememb v podatkih LiDAR. Elektrotehniški vestnik. 84 (3) 2017, 103-107. 54. Seme, S., N. Lukač, B. Štumberger, M. Hadžiselimović. Power quality experimental analysis of grid-connected photovoltaic systems in urban distribution networks. Energy. 139 2017, 1261-1266. 55. Lukač, N., G. Štumberger, B. Žalik. Wind resource assessment using airborne LiDAR data and smoothed particle hydrodynamics. Environmental Modelling & Software. 95 2017, 1-12. 56. Belič, E., N. Lukač, K. Deželak, B. Žalik, G. Štumberger. GPU-based online optimization of low voltage distribution network operation. IEEE transactions on smart grid. 8 (3) 2017, 1460-1468. 57. Horvat, D., B. Žalik. Inclusion test for polyhedra using depth value comparison on the GPU. International journal of computer theory and engineering. 9 (2) 2017, 137-141. 58. Dugonik, B., A. Vučinič Dugonik, D. Horvat, B. Žalik, D. Špelič. e-Derma - a novel wireless dermatoscopy system. Journal of medical systems. 2017, 1-12. 59. Žalik, B., D. Mongus, K. Rizman Žalik, N. Lukač. Boolean operations on rasterized shapes represented by chain codes using space filling curves. Journal of visual communication and image representation. 49 2017, 420-432. 60. Rizman Žalik K., B. Žalik. Multi-objective evolutionary algorithm using problem-specific genetic operators for community detection in networks. Neural computing & applications. 30 (9) 2018, 2907-2920. 61. Kohek, Š., D. Strnad, B. Žalik, S. Kolmanič. Estimation of projection matrices from a sparse set of feature points for 3D tree reconstruction from multiple images. Periodicals of engineering and natural sciences. 5 (3) 2017, 278-285. 62. Markovič, R., J. Peltan, M. Gosak, D. Horvat, B. Žalik, B. Seguy, R. Chauvel, G. Malandain, T. Couffinhal, C. Dupláa, M. Marhl, E. Roux. . Planar cell polarity genes frizzled4 and frizzled6 exert patterning influence on arterial vessel morphogenesis. PloS one. 12 (3) 2017, 1-19. 63. Kolmanič, S., S. Tojnko, T. Unuk, Š. Kohek. The computer-aided teaching of apple tree pruning and training. Computer applications in engineering education. 25 (4) 2017, 568-577. 64. Strnad, D., A. Nerat, Š. Kohek. Neural network models for group behavior prediction : a case of soccer match attendance. Neural computing & applications. 28 (2) 2017, 287-300. 65. Strnad, D., Š. Kohek. Novel discrete differential evolution methods for virtual tree pruning optimization. Soft computing. 21 (4) 2017, 981-993. 66. Žalik, B., D. Mongus, K. Rizman Žalik, N. Lukač. Chain code compression using string transformation techniques. Digital signal processing. 53 2016, 1-10. 67. Srećković, N., G. Štumberger, N. Lukač. Vpliv dodatnih fotonapetostnih sistemov na prihranke energije v srednjenapetostnem distribucijskem omrežju. Elektrotehniški vestnik. 83 (1/2) 2016, 1-5. 127 GeMMA Activity Report 2016–2022 68. Belič, E., K. Dežan, N. Lukač, G. Štumberger. Analiza prihrankov energije v nizkonapetostnem omrežju, doseženih z optimalno energijo jalove moči fotovoltaičnih sistemov. Elektrotehniški vestnik. 83 (1/2) 2016, 37-41. 69. Srećković, N., N. Lukač, B. Žalik, G. Štumberger. Determining roof surfaces suitable for the installation of PV (photovoltaic) systems, based on LiDAR (Light Detection And Ranging) data, pyranometer measurements, and distribution network configuration. Energy. 96 2016, 404-414. 70. Lukač, N., S. Seme, K. Dežan, B. Žalik, G. Štumberger. Economic and environmental assessment of rooftops regarding suitability for photovoltaic systems installation based on remote sensing data. Energy. 107 2016, 854-865. 71. Horvat, D., B. Žalik, D. Mongus. Context-dependent detection of non-linearly distributed points for vegetation classification in airborne LiDAR. ISPRS journal of photogrammetry and remote sensing : official publication of the International Society for Photogrammetry and Remote Sensing. 116 2016, 1-14. 72. Žalik, B., D. Mongus, Y. K. Liu, N. Lukač. Unsigned Manhattan Chain Code. Journal of visual communication and image representation. 38 2016, 186-194. Review Article 1. Rizman Žalik, K. Odkrivanje skupnosti v električnih omrežjih. Elektrotehniški vestnik online. 88 (5) 2021, 273-278. Short Scientific Article 1. Kohek, Š. Interactive synthesis and visualisation of vast areas with geometrically diverse trees. Informatica : an international journal of computing and informatics. 44 (1) 2020, 109-110. Professional Papers 1. Strnad, D. Podatkovna struktura za disjunktne množice. Presek : list za mlade matematike, fizike, astronome in računalnikarje. 49 (1) 2021/2022, 22-26. 2. Šturm, T., R. Pisek, B. Repnik, D. Matijašič. Pregledovalnik podatkov o gozdovih = Forest data viewer. Geodetski vestnik : glasilo Zveze geodetov Slovenije. 61 (1) 2017, 125-131. 3. Strnad, D. Deljenje skrivnosti. Presek : list za mlade matematike, fizike, astronome in računalnikarje. 45 (5) 2017/2018, 26-28. 4. Strnad, D. Drevesno preiskovanje Monte Carlo. Presek : list za mlade matematike, fizike, astronome in računalnikarje. 44 (6) 2016/2017, 21-27. 128 GeMMA Activity Report 2016–2022 5. Triglav Čekada, M., S. Tršan, B. Pegan Žvokelj, N. Lukač, M. Bizjak, M. Brumen, B. Žalik. STEZA - stereozajem iz aerofotografij in podatkov LIDAR = STEZA - combined stereorestitution from aerophotographs and LIDAR data. Geodetski vestnik : glasilo Zveze geodetov Slovenije. 60 (2) 2016, 285-288. Published Scientific Conference Contributions 1. Kavran, D., B. Žalik, N. Lukač. Time series augmentation based on beta-VAE to improve classification performance. In: ROCHA, Ana Paula (ed.), STEELS, Luc (ed.), HERIK, H. Jaap van den (ed.). ICAART 2022, 4th International Conference on Agents and Artificial Intelligence, February, 2022. Online streaming. 2 2022, pp. 15-23. 2. Žnidarič, A., M. Kreslin, A. Anžlin, A. Krivic, D. Mongus. Towards automated detection of cracked concrete. In: PELLEGRINO, Carlo (ed.). Proceedings of the 1st Conference of the European Association on Quality Control of Bridges and Structures : EUROSTRUCT 2021. 2022, pp. 1294-1300. 3. Granda, A., N. Lukač, A. Pur, Š. Kohek, Efficient machine learning based graph layout calculation for investigation platform. In: ČIBEJ, Uroš (ed.). Proceedings of the 8th Student Computing Research Symposium (SCORES22), October, 2022, Ljubljana, Slovenia. 1st ed. Ljubljana: Faculty of Computer and Information Science; Maribor: Faculty of Electrical Engineering and Computer Science; Koper: Faculty of Mathematics, Natural Sciences and Information Technologies. 2022, pp. 5-8. 4. Železnik, L., D. Strnad, B. Žalik, D. Podgorelec. Brezizgubno stiskanje digitalnega avdia s prileganjem daljic in kvadratnih Bézierovih krivulj. In: ŽEMVA, Andrej (ed.), TROST, Andrej (ed.). Proceedings of the 31st International Electrotechnical and Computer Science Conference ERK 2022, September, 2022, Portorož, Slovenia. Ljubljana: Slovenska sekcija IEEE: Fakulteta za elektrotehniko. 2022, pp. 183-186. 5. Horvat, Š., Š. Kohek, D. Ivajnšič, D. Strnad. Detekcija vegetacijskih krajinskih elementov v podatkih LiDAR in ortofoto z nevronsko mrežo. In: ŽEMVA, Andrej (ed.), TROST, Andrej (ed.). Proceedings of the 31st International Electrotechnical and Computer Science Conference ERK 2022, September, 2022, Portorož, Slovenia. Ljubljana: Slovenska sekcija IEEE: Fakulteta za elektrotehniko. 2022, pp. 343-346. 6. Jesenko, D., D. Mongus, U. Lešnik. Influence of COVID-19 On PM[]10PM[]10 Concentrations in Maribor. In: T. Letnik (ed.). 19th European Transport Congress of the EPTS Foundation e. In : European green deal challenges and solutions for mobility and logistics in cities, Conference proceedings, October, 2021, Maribor, Slovenia. 1st ed. Maribor: Zum urbanizem, planiranje, projektiranje: University. 2021, pp. 173-182. 7. Jesenko, D., L. Hruda, I. Kolingerová, B. Žalik, D. Podgorelec. Detection of water levels in lake Cerknica using sentinel-2 data and symmetry. In: S. Y. Yurish (ed.). 3rd International Conference on Advances in Signal Processing and Artificial Intelligence (ASPAI' 2021), Proceedings of the 3rd International Conference on Advances in Signal Processing and Artificial Intelligence (ASPAI' 2021), November 2021, Porto, Portugal. Barcelona: International Frequency Sensor Association (IFSA). cop. 2021, pp. 65-69. 8. Cukjati, J., D. Mongus, B. Žalik. A brief survey on the availability of satellite air pollution data. In: 7th World Multidisciplinary Earth Sciences Symposium (WMESS 2021) September 2021, Prague, Czech Republic, (IOP conference series. Earth and environmental science (Online)), 906 (11) 2021. 9. Podgorelec, D., A. Nerat, B. Žalik. Statistics-based chain code compression with decreased sensitivity to shape artefacts. Informatica : an international journal of computing and informatics, 5th conference MATCOS, October 2019, Koper, Slovenia. 45 (2) 2021, pp. 205-212. 129 GeMMA Activity Report 2016–2022 10. Vlahek, D., T. Stošić, T. Golob, M. Kalc, T. Ličen, M. Vogrin, D. Mongus. Method for estimating tensiomyography parameters from motion capture data. Informatica : an international journal of computing and informatics, 5th conference MATCOS, October 2019, Koper, Slovenia. 45 (2) 2021, pp. 213-222. 11. Kolmanič, S., Š. Kohek, M. Brumen, D. Žlaus, T. Golob, D. Jesenko, D. Mongus. Simulator rasti vegetacije, temelječ na podatkih LiDAR. In: A. Žemva (ed.), A. Trost (ed.). Proceedings of the 30th International Electrotechnical and Computer Science Conference ERK 2021, September 2021, Portorož, Slovenia. Ljubljana: Slovenska sekcija IEEE: Fakulteta za elektrotehniko, 2021, pp. 87-90. 12. Žalik, B., D. Strnad, K. Rizman Žalik, A. Nerat, N. Lukač, B. Lipuš, D. Podgorelec. In: A. Žemva (ed.), A. Trost (ed.). Proceedings of the 30th International Electrotechnical and Computer Science Conference ERK 2021, September 2021, Portorož, Slovenia. Ljubljana: Slovenska sekcija IEEE: Fakulteta za elektrotehniko, 2021, pp. 329-332. 13. Pankelj, M., N. Lukač, S. Rizvić, S. Kolmanič. Ohranjanje kulturne dediščine s pomočjo navidezne in obogatene resničnosti. In: V. Pejović (ed.). Human-Computer Interaction in Information Society, October 2020, Ljubljana, Slovenia. Information Society - IS 2020 : proceedings of the 23rd international multiconference, vol. H. Ljubljana: Institut "Jožef Stefan", 2020, pp. 25-28. 14. Kolmanič, S., M. Marksel, D. Mongus, B. Žalik. Tehnologije navidezne in obogatene resničnosti, kot orodje za predstavitev novih idej in produktov na sejmih : primer MAHEPA. In: Umetna inteligenca - korak k večji uspešnost, 27. konferenca Dnevi slovenske informatike, virtualna konferenca, October, 2020. Ljubljana: Slovensko društvo Informatika, 2020, pp. 109-118. 15. Mongus, D., M. Brumen, B. Kozan. Assessment of lakes' surface elevation from optical satellite images. In: Umetna inteligenca - korak k večji uspešnosti, 27. konferenca Dnevi slovenske informatike, virtualna konferenca, October 2020. Ljubljana: Slovensko društvo Informatika, 2020, pp. 174-182. 16. Kolmanič, S., N. Lukač, S. Rizvić, B. Žalik. Interaktivna razstavna vitrina - orodje za muzejske zbirke prihodnosti. In: A. Žemva (ed.), A. Trostj (ed.). Proceedings of the Twenty-ninth International Electrotechnical and Computer Science Conference ERK 2020, September 2020, Portorož, Slovenia. Ljubljana: Slovenian Section IEEE, 2020, pp. 99-102. 17. Breznar, J., B. Lipuš, S. Kolmanič. Sistem prikazovanja eksponatov muzeja. In: A. Žemva (ed.), A. Trostj (ed.). Proceedings of the Twenty-ninth International Electrotechnical and Computer Science Conference ERK 2020, September 2020, Portorož, Slovenia. Ljubljana: Slovenian Section IEEE, 2020, pp. 331-334. 18. Kavran, D. Time series classification using time-frequency analysis and Convolutional Neural Networks. In: A. Žemva (ed.), A. Trostj (ed.). Proceedings of the Twenty-ninth International Electrotechnical and Computer Science Conference ERK 2020, September 2020, Portorož, Slovenia. Ljubljana: Slovenian Section IEEE, 2020, pp. 444-447. 19. Urh, F., D. Strnad, A. K. Clarke, D. Farina, A. Holobar. On the need for spatial whitening of high-density surface electromyograms in motor unit identification by neural networks. In: T. Jarm (ed.), Proceedings of the EMBEC 2020 : 8th European Medical and Biological Engineering Conference, November, December 2020, Portorož, Slovenia. Cham: Springer Nature. cop. 2021, pp. 915-922. 20. Urh, F., D. Strnad, A. K. Clarke, D. Farina, A. Holobar. On the selection of neural network architecture for supervised motor unit identification from high-density surface EMG*. In: 42nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE. cop. 2020, pp. 736-739. 21. Srećković, N., N. Lukač, B. Žalik, G. Štumberger. Minimization of distribution network losses using optimal PV installation, reactive power generation and network reconfiguration. In: S. Seme (ed.), M. Hadžiselimović (ed.), B. Štumberger (ed.). Conference proceedings, 7th Symposium on Applied Electromagnetics SAEM'2018, June 2018, Podčetrtek, Slovenia. Maribor: University of Maribor Press. 2019, pp. 216-222. 130 GeMMA Activity Report 2016–2022 22. Jesenko, D., D. Mongus, S. Liu, M. Triglav Čekada. Detection of snow levels in the Slovenian Alps at different seasons using Sentinel-1. In: ICESI 2019, Proceedings, International conference on engineering, science, and industrial applications, August 2019, Tokyo, Japan. 2019, pp. 76-79. 23. Lukač, N., B. Žalik. Landslide simulation by using the SPH method and LiDAR data. In: ICESI 2019, Proceedings, International conference on engineering, science, and industrial applications, August 2019, Tokyo, Japan. 2019, pp. 86-94. 24. Urh, F., A. Holobar, D. Strnad. Mutual comparison of different neural networks in identification of motor unit firings from high-density surface electromyograms. In: Summer school on neurorehabilitation, SSNR2019, September 2019, Baiona, Spain. 2019, pp. 1-2. 25. Bizjak, M., G. Štumberger, B. Žalik, N. Lukač. Time series prediction for EMS with machine learning. In: ICESI 2019, Proceedings, International conference on engineering, science, and industrial applications, August 2019, Tokyo, Japan. 2019, pp. 90-94. 26. Srećković, N., N. Lukač, G. Štumberger. Impact of OLTC equipped transformer operation on PV installation in urban distribution network. In: ICREPQ'19, International Conference on Renewable Energies and Power Quality ), April 2019, Tenerife, Canary Islands. 2019, pp. 314-319. 27. Kavran, D., R. Novak, J. Banko, R. Potočnik, L. Pečnik, B. Bošković. Nadgradnja algoritma FLORS za besednovrstno označevanje slovenskih besedil. In: I. Fister (ed.), Proceedings of the 2019 6th Student Computer Science Research Conference - StuCoSReC. Koper: University of Primorska Press, 2019, pp. 91-99. 28. Mongus, D., S. Jurič. Generation of traversability maps based on 3D point-clouds. In: Second International conference on Next generation computing applications 2019. IEEE. cop. 2019, pp. 1-5. 29. Jeromel, A., M. Žalik, M. Brumen, N. Lukač. Visualization of 3D Earth using GIS services. In: A. Brodnik (ed.), G. Galamboš (ed.), B. Kavšek (ed.). Middle-European Conference on Applied Theoretical Computer Science, Proceedings of the 22nd International Multiconference Information Society, October, 2019, Ljubljana, Slovenia, vol. I. Ljubljana: Institut "Jožef Stefan", 2019, pp. 47-50. 30. Kohek, Š., S. Kolmanič, B. Žalik, J. Pihler, R. Tomažič, D. Strnad. Technologies and approaches for effective information-based vegetation management. In: 4. slovenska konferenca o vzdrževanju v elektroenergetiki, November 2018, Nova Gorica, Slovenia. Ljubljana: Slovensko združenje elektroenergetikov CIGRE - CIRED, 2018, pp. 1-7. 31. Colnarič, M., S. Moraus, M. Zorman, G. Žlahtič, J. Završnik, H. Blažun Vošner, M. Turčin, T. Završnik, S. Jurič, B. Slemnik, J. Detela. Pametno okolje za učinkovito ščetkanje zob. In: M. Gams (ed.), A. Tavčar (ed.). Workshop Electronic and mobile health and smart cities, Proceedings of the 21st International Multiconference Information Society, October 2018, Ljubljana, Slovenia, vol. I. Ljubljana: Institut "Jožef Stefan", 2018, pp. 37-39. 32. Kolednik, D., D. Boldin, M. Puhar, B. Repnik, D. Žganec, B. Žalik. Inspire sistem registrov in vzpostavitev storitve slovenskega sistema registrov. In: Š. Urh Popovič (ed.). Digitalizacija in mi. Ljubljana: Slovensko društvo Informatika, 2018, pp. 1-7. 33. Žlaus, D., D. Mongus. In-place SIMD accelerated mathematical morphology. In: 2018 International Conference on Big Data and Education -ICBDE 2018, March 2018, Honolulu, Hawaii, USA. New York: ACM. cop. 2018, pp. 1-5. 34. Jesenko, D., M. Brumen, N. Lukač, B. Žalik, D. Mongus. Visualization and analytics tool for multi-dimensional data. In: 2018 International Conference on Big Data and Education - ICBDE 2018, March 2018, Honolulu, Hawaii, USA. New York: ACM. cop. 2018, pp. 1-5. 35. Jeromel, A. Real-time visualization of 3D digital Earth. In: I. Fister (ed.), A. Brodnik (ed.). Proceedings of the 2018 5th Student Computer Science Research Conference - StuCoSReC. Koper: University of Primorska Press, 2018, pp. 39-42. 131 GeMMA Activity Report 2016–2022 36. Kavran, D. Time series classification with Bag-Of-Words approach. In: I. Fister (ed.), A. Brodnik (ed.). Proceedings of the 2018 5th Student Computer Science Research Conference - StuCoSReC. Koper: University of Primorska Press, 2018, pp. 55-59. 37. Vlahek, D. The problem of quantum computers in cryptography and post-quantum cryptography. In: I. Fister (ed.), A. Brodnik (ed.). Proceedings of the 2018 5th Student Computer Science Research Conference - StuCoSReC. Koper: University of Primorska Press, 2018, pp. 61-64. 38. Kohek, Š., D. Strnad, B. Žalik, S. Kolmanič. Estimation of projection matrices from a sparse set of feature points for 3D tree reconstruction from multiple images. In: O. Çinar (ed.). Book of proceedings, 2017, pp. 171-178. 39. Strnad, D., Š. Kohek, S. Kolmanič, B. Žalik. A feedback loop model of interaction between soil characteristics and vegetation in afforestation simulator ForestMAS. In: O. Çinar (ed.). Book of proceedings, 2017, pp. 290-293. 40. Jesenko, D., N. Lukač, M. Bizjak, B. Žalik. Parallelization of a two-level evolutionary algorithm for discovering relations between nodes' features in a complex network using GPGPU. In: Conference proceeding, 7th international conference of engineering and applied sciences - ICEAS 2017, June 2017, Toronto, ON, Canada. SEAS International. cop. 2017, pp. 8-13. 41. Lukač, N., D. Jesenko, M. Bizjak, B. Žalik. GPU-based DBSCAN clustering on locality sensitive hashing. In: Conference proceeding, 7th international conference of engineering and applied sciences - ICEAS 2017, June 2017, Toronto, ON, Canada. SEAS International. cop. 2017, pp. 14-19. 42. Špelič, D., D. Podgorelec, T. Kajtna. MA reaction- mobile application for measuring and training of reaction time. In: Conference proceeding, 7th international conference of engineering and applied sciences - ICEAS 2017, June 2017, Toronto, ON, Canada. SEAS International. cop. 2017, pp. 28-36. 43. Bizjak, M., N. Lukač, D. Jesenko, B. Žalik. Estimating boundaries of an object's faces in unstructured 3D point clouds. In: Conference proceeding, 7th international conference of engineering and applied sciences - ICEAS 2017, June 2017, Toronto, ON, Canada. SEAS International. cop. 2017, pp. 88-94. 44. Lukač, N., B. Žalik, G. Štumberger. Photovoltaic potential assessment and ranking of rooftops segments based on LiDAR data. In: ICREPQ'17, International Conference on Renewable Energies and Power Quality, April 2017, Malaga, Spain. 2017, pp. 1-5. 45. Hämmerle, M., N. Lukač, K.-C. Chen, Z. Koma, C.-K. Wang, K. Anders, B. Höfle. Simulating various terrestrial and UAV lidar scanning configurations for understory forest structure modelling. In: D. Li (ed.). ISPRS Geospatial Week 2017, September 2017, Wuhan, China. Vol. IV-2/W4. ISPRS, 2017, pp. 59-65. 46. Lipuš, B., B. Žalik. Evaluation of the marker radius on the efficiency of terrestrial Lidar data watermarking. In: Photogrametry and remote sensing cartography and GIS, Conference proceedings, vol.17. 17th International Multidisciplinary Scientific Geoconference, SGEM 2017, June, July 2017, Albena, Bulgaria. Sofia: STEF92 Technology, 2017, pp. 111-118. 47. Fister I., N. Lukač. First three years of the ACM student chapter Maribor. In: I. Fister (ed.), A. Brodnik (ed.). Proceedings of the 2017 4th Student Computer Science Research Conference - StuCoSReC. Koper: University of Primorska Press, 2017, pp. 13-14. 48. Jesenko, D., U. Vilhar, M. Skudnik, B. Žalik, D. Mongus. Determination of the threshold function for defining a complex network of tree competition with an evolutionary algorithm. In: A. Žemva (ed.), A. Trost (ed.). Proceedings of the Twenty-sixth International Electrotechnical and Computer Science Conference ERK 2017, September 2017, Portorož, Slovenia. Ljubljana: IEEE, Slovenska sekcija IEEE, 2017, pp. 283-286. 49. Žlaus, D., D. Mongus. Detection of regions of changes of Earths' surface using SAR imagery. In: A. Žemva (ed.), A. Trost (ed.). Proceedings of the Twenty-sixth International Electrotechnical and Computer Science Conference ERK 2017, September 2017, Portorož, Slovenia. Ljubljana: IEEE, Slovenska sekcija IEEE, 2017, pp. 399-402. 132 GeMMA Activity Report 2016–2022 50. Kolmanič, S., Š. Kohek. Gap regeneration simulation employing ellenberg ecological values and realistic real-time forest visualization. In: O. Çinar (ed.). Book of proceedings, 2017, pp. 294-300. 51. Jesenko, D. A novel prediction method based on polynomial fitting. In: The 11th International Conference on Information Technology and Applications. Piscataway: IEEE. cop. 2016, pp. 1-4. 52. Rizman Žalik, K. Multi-objective community detection method using an improved NSGA II algorithm. In: HAMZA, H. Mohamed (ed.). Conference proceedings, The 14th IASTED International Conference on Software Engineering - SE 2016, February 2016, Innsbruck, Austria. Calgary: IASTED, 2016, pp. 327-334. 53. Horvat D., D. Kobale, J. Zorec, D. Mongus, B. Žalik. Assessing the usability of LiDAR processing methods on UAV Data. In: H. R. Arabnia (ed.), L. Deligiannidis (ed.), M. Yang (ed.). CSCI 2016 : Proceedings, 2016 International conference on computational science and computational intelligence, December 2016, Las vegas, Nevada, USA. Los Alamitos (California); Tokyo: IEEE: CPS. cop. 2016, pp. 665-670. 54. Bizjak, M. The segmentation of a point cloud using locally fitted surfaces. In: E. Kyriakides (ed.). Proceedings of the 18th Mediterranean electromechanical conference, Intelligent & efficient etchnologies & services for the citizen. IEEE. cop. 2016, pp. 1-6. 55. Mongus, D. Izločanje informacij iz 3D slik površja. In: B. Potočnik (ed.). ROSUS 2016 : Računalniška obdelava slik in njena uporaba v Sloveniji 2016, Zbornik 11. strokovne konference, March 2016, Maribor, Slovenia. Maribor: Fakulteta za elektrotehniko, računalništvo in informatiko, Inštitut za računalništvo, Laboratorij za sistemsko programsko opremo. 2016, pp. 12-21. 56. Kohek, Š., D. Horvat, D. Strnad, B. Žalik. Paralelizacija razporeditve virov pri simulaciji rasti dreves na grafični procesni enoti s stalnimi nitmi. In: B. Zajc (ed.), A. Trost (ed.). Zbornik petindvajsete mednarodne Elektrotehniške in računalniške konference ERK 2016, September 2016, Portorož, Slovenia. Ljubljana: IEEE Region 8, Slovenska sekcija IEEE. 2016, vol. B, pp. 19-22. 57. Podgorelec, D., A. Nerat, D. Špelič. Zlivanje dveh urejenih zaporedij na mestu v linearnem času. In: B. Zajc (ed.), A. Trost (ed.). Zbornik petindvajsete mednarodne Elektrotehniške in računalniške konference ERK 2016, September 2016, Portorož, Slovenia. Ljubljana: IEEE Region 8, Slovenska sekcija IEEE. 2016, vol. B, pp. 27-30. 58. Lukač, N., D. Kolednik, M. Bizjak, D. Strnad, G. Štumberger, B. Žalik. Razpoznava in analiza fotovoltaičnih sistemov s podatki daljinskega zaznavanja. In: B. Zajc (ed), A. Trost (ed.). Zbornik petindvajsete mednarodne Elektrotehniške in računalniške konference ERK 2016, September 2016, Portorož, Slovenia. Ljubljana: IEEE Region 8, Slovenska sekcija IEEE. 2016, vol. B, pp. 67-70. 59. Kolmanič, S., R. Cvirn, D. Jesenko, B. Lipuš. Simulacija dinamike vrzeli s simulatorjem ForestMAS. In: B. Zajc (ed.), A. Trost (ed.). Zbornik petindvajsete mednarodne Elektrotehniške in računalniške konference ERK 2016, September 2016, Portorož, Slovenia. Ljubljana: IEEE Region 8, Slovenska sekcija IEEE. 2016, vol. B, pp. 71-74. Published Professional Conference Contributions 1. Mongus, D. Achievements of the candidate. In: ŽALIK, Borut (ed.), GAMS, Matjaž (ed.). Legends of Computing and Informatics, Information Society-IS 2022. Proceedings of the 25th International Multiconference , volume J. October, 2022. Ljubljana: Institut "Jožef Stefan". 2022, pp. 21-23. 133 GeMMA Activity Report 2016–2022 2. Repnik, B., D. Žganec, D. Mongus, B. Žalik. Integracija senzorskih tokov za izboljšanje požarne varnosti. In: Š. Urh Popovič (ed.). Ustvarjamo prihodnost priložnosti. Ljubljana: Slovensko društvo Informatika, 2017, pp. 1-6. 3. Pevec, A., B. Repnik, M. Puhar, M. Pegan, D. Žganec, B. Žalik. Nadgradnja metapodatkovnega sistema za prostorsko referencirane vire, temelječega na aplikaciji GeoNetwork. In: Š. Urh Popovič (ed.). Ustvarjamo prihodnost priložnosti. Ljubljana: Slovensko društvo Informatika, 2017, pp. 1-10. Independent Scientific Component Parts or Chapters in a Monography 1. Novak, N., Jesenko, D., Drešček, U., Triglav Čekada, M. Zbiranje prostovoljnih fotografij topografskih sprememb ter ocena njihove uporabnosti. In: BREG VALJAVEC, Mateja (ed.). Preteklost in prihodnost. Ljubljana: Založba ZRC. 2022, pp. 321-332. 2. Lukač, N., B. Žalik, G. Štumberger. Segmentation of 3D point cloud data for assessment and ranking of photovoltaic potential. In: M. Pérez-Donsión (ed.), G. Vitale (ed.). Advances in renewable energies and power quality. Cambridge: Cambridge Scholars Publishing. cop. 2020, pp. 27-36. 3. Pavlič, L., S. Rizvić, D. Mongus. Digital production pipeline for virtual cultural heritage applications using interactive storytelling. In: A. Lugmayr (ed.). Information systems and management in media and entertainment industries. Cham: Springer. cop. 2016, pp. 223-244. Patent Applications 1. Žalik, B., B. Lipuš, A. Nerat, D. Podgorelec. Procedure for detecting the reflectional symmetry of geometrical objects, application number EP 21217758.8, 24 December 2021. EU Patent, 2021. 20 pages. 2. Lipuš, B., B. Žalik. Postopek označevanja velikega nestrukturiranega oblaka tridimenzionalnih točk z digitalnim vodnim tiskom, application number P-201600316, 30 December 2016. Ljubljana: Urad RS za intelektualno lastnino, 2016. 7 pages. Patents 1. Lukač, N., B. Žalik. Method and apparatus for near-lossless compression and decompression of 3D meshes and point clouds, Application number US9734595 (B2), filed 15 August 2017. United States Patent Office, 2017. 55 pages. 2. Žalik B., D. Mongus. Light detection and ranging (Lidar)data compression and decompression methods and apparatus, Application number 13/289.839, filed 4 November 2011. United States Patent and trademark office No.: US 9.300.321 B2, date of patent 29 March 2016. 15 pages. 134 GeMMA Activity Report 2016–2022 Editorial Boards Membership 1. Data in brief. New York: Elsevier, 2014-. ISSN 2352-3409. Lukač, N. (editor 2018-2022). 2. Land. Basel: MDPI AG, 2012. ISSN 2073-445X. Bizjak, M. (guest editor 2022) 3. Sensors. Basel: MDPI, 2001. ISSN 1424-8220. Rizman Žalik, K. (guest editor 2022) 4. Energies. Basel: Molecular Diversity Preservation International, 2008-. ISSN 1996-1073. Lukač, N. (guest editor 2019) 5. ISPRS international journal of geo-information. Basel: MDPI, 2012-. ISSN 2220-9964. Bizjak, M. (guest editor 2021), Lukač, N. (editor 2020-2021, guest editor 2021), Mongus, Domen (editor 2019-2020). 6. Sensors. Basel: MDPI, 2001-. ISSN 1424-8220. Mongus, D. (editor 2019). 7. IKT sistem za optimizacijo dostav in tovornega prometa Dravinjske doline, končno poročilo o izvedbi projekta. Maribor, Univerza v Mariboru, 2018. M. Mencinger, A. Soderžnik, M. Garmut, L. Sreš, J. Jelenc, M. Fale, D. Mlinarič, S. Grad, I. Štampar, J. Kukovič, S. Božičnik (editor), D. Mongus (editor), T. Letnk (editor). 8. Dostave TDD - Organiziranje in optimiranje dostav tovora na območju Dravinjske doline, končno poročilo o izvedbi projekta. Maribor, Univerza v Mariboru, 2017. M. Mencinger, U. Červan, V. Lorenčič, M. Bogolin, S. Plantak, M. Sovič, G. Vogrin, R. Koletnik, L. Vidrač, T. Šklebek, A. Želj, S. Božičnik (editor), I. Peruš (editor), D. Mongus (editor), T. Letnik (editor). 9. GeMMA 2000-2016: from fundamental geometric algorithms to advanced methodologies for pattern recognition and dynamics modelling in large earth observation datasets. Maribor: Faculty of Electrical Engineering and Computer Science, Laboratory for Geometric Modelling and Multimedia Algorithms, 2016. D. Podgorelec (editor). 10. Human-Computer Interaction in Information Society. Information Society - IS 2020, Proceedings of the 23rd international multiconference, October 2020, Ljubljana, Slovenia. Vol. H. Ljubljana: Institut "Jožef Stefan", 2020. V. Pejović (editor), M. Kljun (editor), V. Groznik (editor), D. Šoberl (editor), K. Čopič Pucihar (editor), B. Blažica (editor), J. Žabkar (editor), M. Pesek (editor), J. Guna (editor), S. Kolmanič (editor). 135 GeMMA Activity Report 2016–2022 Prizes, Awards, Honours and Medals UM Awards and Honours for Employees 2022: Štefan Kohek, Recognition for exceptional contributions to scientific and pedagogical reputation and excellence of University of Maribor 2019: Niko Lukač, Recognition for exceptional contributions to scientific and pedagogical reputation and excellence of University of Maribor 2018: Domen Mongus, Award for exceptional contributions to scientific and pedagogical reputation and excellence of University of Maribor UM FERI Awards and Honours for Employees 2022: Dino Vlahek, Award for exceptional research achievements 2022: Matej Brumen, Plaque FERI for research and development work 2021: Blaž Repnik, Plaque FERI for research and development work 2019: Domen Mongus, Award for exceptional research achievements 2018: Damjan Strnad, Simon Kolmanič, Niko Lukač, Plaque FERI for professional work UM Award for Research Work of Students (Andrej Perlach’s Award) 2021: Aljaž Jeromel 136 GeMMA Activity Report 2016–2022 UM FERI Awards for Students Mitja Žalik, Plaque for the Best Graduate of the Second Cycle of University Study Programmes (2022) Aljaž Žel, Award for Exceptional Student Contribution (2022 – UPM, 2022 – IEEExtreme) Mitja Žalik, Award for Exceptional Student Contribution (2018, 2019, 2020, 2021 – UPM, 2021 – IEEExtreme) Mitja Žalik, Plaque for the Best Graduate of the First Cycle of University Study Programmes (2020) 137 GeMMA Activity Report 2016–2022 Ph.D. Candidates Granted by ARRS and Completed Ph.D.s Supervised in GeMMA Ph.D. Candidates granted by Slovene Research Agency Ph.D. candidates supervised by professor dr. Borut Žalik: • Niko Lukač, 1 December 2012 – 31 May 2016; • Denis Horvat, 1 October 2013 – 31 March 2017; • David Jesenko, 1 November 2014 – 30 April 2018; • Jernej Cukjati, 1 October 2019 – 30 September 2023. Ph.D. candidates supervised by associate professor dr. Domen Mongus: • Danijel Žlaus, 1 October 2016 – 22 March 2021. Ph.D. candidates supervised by associate professor dr. Niko Lukač: • Niko Uremović, since 1 October 2022 . Completed Ph.D. studies in Computer Science All 5 disertations have been completed through UM FERI computer science & informatics study programme. Two of the disertations have been supervised by professor dr. Borut Žalik, and other three by associate professor dr. Domen Mongus, associate professor dr. Damjan Strnad, and assistant professor dr. Niko Lukač. • Jesenko, David. Algoritem določanja funkcijske odvisnosti povezav med vozlišči v kompleksnih mrežah (An algorithm for determining the functional relation of nodes' conectivity in complex networks), 2018; • Lukač, Niko. Algoritem za celostno vrednotenje fotovoltaičnega in vetrnega potenciala večjih geografskih območij (Algorithm for the determination of photovoltaic and wind pontential over large geographic areas), 2016; • Žlaus, Danijel. Algoritem za učinkovit izračun verige elementarnih morfoloških filtrov na centralni procesni enoti (Algorithm for efficient computation of elementary morphological filter chain on central processing unit), 2021; 138 GeMMA Activity Report 2016–2022 • Kohek, Štefan. Interaktivna tvorba in prikaz obsežnih področij geometrijsko raznolikih dreves (Interactive synthesis and visualization of vast areas gemoetrically diverse trees), 2019; • Bizjak, Marko. Algoritem za napovedovanje toplotne obremenitve stavb na večjem geografskem območju (An algorithm for estimating thermal load of buildings on a large geographic area), 2019; 139 GeMMA Activity Report 2016–2022 Functions and Honours in National and International Associations dr. Borut Žalik: • Senior member of Association for Computing Machiner (ACM). ACM brings together computing educators, researchers, and professionals to inspire dialogue, share resources, and address the field's challenges. It has nearly 100,000 members around the globe and it has led to Councils in Europe, India, and China, with its growing membership, fostering networking opportunities that strengthen ties within and across countries and technical communities. • Member of European Academy of Sciences and Arts (EASA) since 2014. The European Academy of Sciences and Arts is a non-governmental, European association committed to promoting scientific and societal progress. Founded in 1990 as a learned society, its 2,000 members are leading scientists, artists, and practitioners of governance, who are dedicated to innovative research, interdisciplinary and transnational collaboration as well as the exchange and dissemination of knowledge. Academy members are elected for their outstanding achievements in science, arts, and governance. dr. Domen Mongus: • Member of Executive Committee of European Umbrella Organisation for Geographic Information (EUROGI) from 2013 to 2019. EUROGI was established in 1994 by the European Commission with the mission is to maximise the availability, effective use and exploitation of geographic information throughout Europe in order to ensure good governance, economic and social development, environmental protection and sustainability, and informed public participation. Within the EUROGI’s ExCom, Domen Mongus was serving two terms as a project portfolio leader in charge of coordinating project activities of the organisation. He was a coordinator of organisation’s policy positions in regards to Big Data and Internet of Things. • Member of Executive Committee of Association Operating in the Field of Geographical Information Systems (GISIG) from 2019 to present. GISIG is an Association operating in the field of Geographical Information Systems, grouping organisations from more than 20 European Countries. It represents a reference centre for common initiatives among Geographical Information Systems operators and users, also acting through the promotion of European projects and the establishment of thematic networks and national secretariats. 140 GeMMA Activity Report 2016–2022 dr. Niko Lukač: • Expert Evaluator for Horizon Europe Framework Programme in 2022. Niko Lukač was an expert evaluator under European Commission for Innovation Action type of project calls within the Horizon Europe Framework Programme in 2022. The calls topic was related to the New Horizon Destinations for Efficient, sustainable and inclusive energy use, and Highly energy-efficient and climate neutral European building stock. • Member of Executive Committee of European Umbrella Organisation for Geographic Information (EUROGI) from 2019 to 2022. Niko Lukač took the position as the ExCom from Domen Mongus in 2019, where we continued his work. • Section Editor for Data in Brief Journal from 2019 to 2022 Niko Lukač is a Section Editor for the field of Computer Science for the rapidly growing journal Data in brief, an open access journal by Elsevier in order to publish complementary papers for promoting and spreading scientific datasets. The journal is being managed by 2 Editors in Chief, 23 Section Editors, and 128 Editors of Editorial Board. The journal received more than 3000 submissions in 2019 alone, and almost twice the number of submissions are expected in 2020. 141 GeMMA Activity Report 2016–2022 Tamara Golob Univerza v Mariboru, Fakulteta za elektrotehniko, računalništvo in informatiko, Maribor, Slovenija t.golob@um.si Abstract: The primary aim of this publication is to support the dissemination of the research achievements of GeMMA Laboratory at the Faculty of Electrical Engineering and Computer Science of the University of Maribor. It follows the first survey published in 2016, and therefore, the actual book concentrates on the research results since then. The previous book, covering a period of 17 years, presented 35 R&D projects, while the new review of activities over the last 7 years covers as many as 58 projects. This growth is a good cue to introduce the secondary, equally important aim of the book. Namely, we would like to leave the track to our successors to stimulate them for even better research results, to show them, what is possible to achieve in 22 years starting from scratch with the will, hard work, orientation towards the applications, devotion to the research work, and the strong team spirit. Keywords: research & development projects, geospatial modelling, multimedia, artificial intellingence, algorithms, industrial cooperation, international cooperation DOI https://doi.org/10.18690/um.feri.2.2023 ISBN 978-961-286-713-3 Institute of Computer Science eMA