Joep Crompvoets
Public Governance Institute, KU Leuven, Belgium (joep.crompvoets@kuleuven.be)
GeoAI at National Mapping Agencies in Europe
Abstract: The adoption of Artificial Intelligence (AI) at National Mapping and Cadastral Agencies (NMCAs) in Europe could bring benefits such as personalised services for citizens, fewer repetitive tasks for staff
members (liberating time for more tasks that have more value as listening, creating, improving services, …), detect harmful content, support decision-making processes where a high number of data and variables are involved, which is more and more frequent when societal challenges are involved.
In this context, GeoAI is defined as AI-applications at NMCAs dealing with geospatial issues. However there has been limited initiatives aiming at measuring and understanding the level of AI adoption by NMCAs in Europe. As a follow up action of the successful Joint Virtual Workshop organised by EuroSDR in collaboration with EuroGeographics on“Artificial Intelligence for NMCAs” held on 3-4 April 2021, the Executive Management Team of EuroSDR decided to repeat the EuroSDR survey on Machine Learning / Deep Learning conducted in 2018. EuroSDR received valuable support of swisstopo to successfully relaunch this survey. The results were presented and discussed during a new EuroSDR/EuroGeographics virtual workshop with around ±150 participants on 27-28 October 2022. 38 mapping agencies in 20 countries (Austria, Belgium, Denmark, Estonia, Finland, France, Greece, Ireland, Italy, Latvia, Netherlands, Norway, Portugal, Romania, Slovenia, Spain, Sweden, Switzerland, Turkey, United Kingdom) responded to the survey. Although this is one of the first attempts to draw the landscape of the use of AI at NMCAs in Europe, the survey provides a unique overview of the current use of AI at NMCAs in Europe – insights which are extremely lacking at this stage and some recommendations to support AI adoption in order to improve their services.
The main outcomes of the survey are that AI is an opportunity for NMCAs (e.g. cost/time savings, more efficient business processes), and AI is not considered to be a threat for NMCAs (although there some concerns about ethics and overuse). In general, the AI-Maturity at NMCAs in Europe is rather low as many applications are still in the stage of Proof-of-Concept – however, we recently see more successful applications being implemented at NMCAs. Besides the survey results, the main results of the workshop polls will also be presented and discussed
Bio: Prof. dr. ir. Joep Crompvoets is full professor and research manager on ‘Information Management in the Public Sector’ at the Public Governance Institute in KU Leuven. He is also secretary-general of EuroSDR – an European Spatial Data Research network linking national mapping and cadastral agencies with academic institutions across Europe for doing applied research. In addition, he is founder of the joint European master programme ‘PIONEER – Public Sector Innovation and E-Governance’. His main research topics refer to (spatial) data infrastructures, GIS, smart cities, governance, and digital transformation. Finally, he has been involved in numerous (inter)national projects somehow related to these research topics.
Glenn Vancauwenberghe
Division SADL, KU Leuven, Belgium (glenn.vancauwenberghe@kuleuven.be)
Co-creating SDI and EO Education for the Future: Experiences and results of the SEED4NA project
Abstract: Initiatives to promote and coordinate the sharing of geospatial and earth observation data have emerged at local, national, regional and international levels. To support the establishment of these data infrastructures but also the uptake and use of the data they provide, specific skills sets are in demand. For SDI and EO education to be effective, it should allow students and future professionals to acquire the right skills on GI, SDI, EO and related technologies. Effective teaching and learning also demands for active engagement of teachers, students and practitioners, each contributing their own skills and knowledge and utilizing their unique background and domain knowledge. Co-creation in education supports cross-fertilization and mutual learning among different stakeholders, allowing them to come to new insights and innovative solutions together. It is widely recognized that teaching should not be a one-way process flowing from teacher to student, since students construct knowledge and meaning from lived experiences rather than from passively taking in information. Co-creation in education changes the dynamic with students no longer being perceived as passive consumers but as active participants. The engagement of professionals in the co-creation process allows to put emphasis on real-world challenges and cases, and better address the needs of society. In this presentation we will introduce co-creation as a new approach to teaching and learning on SDI and earth observation. We will discuss how co-creation was applied in the SEED4NA project on SDI/EO Education and Training for North Africa, and reflect on more innovative ways of co-creating SDI and EO education for the future.
Bio: Dr. Glenn Vancauwenberghe is a Research Manager at the Spatial Applications Division (SADL), KU Leuven, Belgium. His main field of research is related to the governance and performance of inter-organizational information sharing. He obtained an MSc degree in Sociology, a MSc degree in the Management of Public Organizations, and a PhD degree in Social Sciences. He has been involved as principal investigator in several projects in the domain of spatial data and geo-information in Europe, in which he mainly focused on governance and data policies, the integration of spatial data in e-government and the performance of spatial data policies. Dr. Glenn Vancauwenberghe is the coordinator of the Erasmus+ SEED4NA (Spatial Data Infrastructure and Earth Observation Education and Training for North Africa) project approved for the period 2020-2023.
Habib Mahi
Habib Mahi, Director of CTS/ASAL, Algeria (hmahi@cts.asal.dz)
Earth Observation in the National Space Program
Abstract – In this talk, we will expose first the missions, organization and Earth Observation Space Algerian Systems and in the second part, we present some thematic applications conducted by Algerian Space Agency for the benefit of national partners.
Bio: Habib Mahi received his Phd degree in computer science from the University of Mohamed Boudiaf in Algeria in 2014 and his MS degree in Techniques Spatiales et Applications from the Centre des Techniques Spatiales in 2004. He is currently the head of the Centre des Techniques Spatiales of the Algerian Space Agency and a permanent researcher in Earth Observation Department. His research interests include: image feature extraction and classification algorithms development for remotely sensed data.
Rolf de By
Department of Geo-information Processing, Digital Society Institute, Faculty of Geo-Information Science and Earth Observation, University of Twente, Netherlands (https://research.utwente.nl/en/persons/rolf-de-by)
How to do SDI when and where there is none
Abstract: The benefits of established (N)SDI are well-documented. SDI helps mature national awareness of resources, resilience, risks and vulnerability. SDI helps in informing better the private sector, and thus sheds a light on investment opportunity as it lowers the associated risks. SDI also helps to inform the governance sector as the intelligence that it offers is an important instrument in evidence-based policy definition. SDI tells us what is where and tells us what is not there. As with most types of infrastructure, the benefits of SDI can only be reaped after its realization. One must lose a fly to catch a trout. It is no coincidence that quite some higher economies have succeeded in constructing successful SDIs, while quite some middle and lower economies have not. We shall look into those reasons and try understand which are the required conditions for setting up a successful (N)SDI, and what can be done when such conditions are insufficiently met.
In the second part of our discussion, we will look at a notion of `SDI light’, distinguished from the above `SDI heavy’ by its lack of institutional custodian anchor and some ad hoc characteristics that cause it to be more vulnerable over time.We will argue that the true challenge is in bootstrapping SDI in such a way that the early investments are low and low-risk and that evolutionary steps in SDI development need to plateau rapidly and frequently. Like so many information systems generally, SDI needs to become `fit in context’ which means that it should provide maximal functionality to what is at first a minimal clientele. As an SDI is made to evolve, tactical choices are required at different fronts: use cases and client needs, resource specificity versus genericity, early and later business cases, governmental acceptance and the position of mandated agencies. All of this may be perceived to sketch a highly intricate and complex puzzle to solve, but in the final part of the discussion we will look at a number of recent developments that may give guidance to the evolution of SDI, whether light or heavy.
Bio: Rolf De By is an associate professor in geo-infomation processing at the faculty ITC of the University of Twente. He obtained his engineer’s (ir.) degree in applied mathematics, and his doctorate (dr.) degree in computer science in the same university, respectively in 1985 and 1991. Originally, his research developed around database technology, database design and collaborative transactional systems, with a specific focus on formal methods for computing, specification and information system design. He joined ITC in 1996, and from that moment his research focus gradually evolved towards geospatial information systems and spatial database technology, and their applications. In recent years, application areas that received special attention have been tropical/smallholder agriculture and the associated data acquisition technology and food distribution systems, and land administration systems of developing economies.
Lucas De Oto
Faculty of Geo-Information Science and Earth Observation, University of Twente, Netherlands (l.h.deoto@utwente.nl)
Teaching GI today: an update on pedagogical approaches and instructional design
Abstract: In recent years, the combination of technological advances and spatial data abundance revolutionized the field of geoinformation (GI). New methodologies and techniques established in other fields of knowledge proved to be relevant to keep up to date and fully benefit from all this technological richness. Consequently, new areas of knowledge have emerged, such as geospatial artificial intelligence (GeoAI) or big geodata.
Simultaneously, the establishment in 2015 of the 2030 Agenda for Sustainable Development and its multiple goals by the United Nations (UN), impose a specific framework for the application and further development of geoinformation science. Lastly, the recent COVID-19 pandemic accelerated the transition towards different modalities of distance education as well as the arrival of multiple digital instruments to fulfill this purpose.
All these factors have profoundly impacted the way of teaching with GI and about GI. Both technical and socio-emotional skills required to successfully perform as a geoscientist in the near future are changing. And so are the means to learn those skills. As a result, the training curriculum for educators in this field is being revised and updated. In this presentation, we will first discuss the challenges currently faced by educators in the field of geosciences and explore new didactic and pedagogical proposals to overcome them. We will then review how to successfully implement these new approaches in our daily teaching using the ADDIE (i.e., Analyze-Design-Develop-Implement-Evaluate) model of instructional design.
Bio: Lucas De Oto graduated as a geographer and lecturer in geography from the University of El Salvador in Buenos Aires, Argentina. He received his MSc degree in Geo-Information for Natural Resources Management at ITC, University of Twente, The Netherlands. In Argentina, he worked both as a consultant geographer and GIS specialist for various ministries and other governmental agencies, notably the National Institute of Geography and the National Institute of Statistics and Censuses. He also collaborated as an adjunct researcher with the National Scientific and Technical Research Council (CONICET) and the Pedagogical University of Buenos Aires (UNIPE). In 2017 he joined the University of Twente where he teaches and coordinates different modules of the master’s program and other courses at the Faculty of Geo-information Science and Earth Observation (ITC). He also collaborates in the development of didactic material, organization, and delivery of international trainings as part of capacity development efforts of the faculty. His current research focusses on the integration of machine learning techniques for spatio-temporal regionalization purposes.
Zvonimir Nevistić
Faculty of Geodesy, University of Zagreb, Croatia (znevistic@geof.hr)
UAV – Flight planning, regulation, processing and applications
Abstract: Unmanned aerial vehicle (UAV) systems are nowadays widely used, especially in applications where aerial imaging is an important source of information for performing certain tasks. With the advancement of technology, UAVs have reached an enviable level of practical reliability and functionality that has enabled these systems to enter the geomatics business market as valuable additional platform for collecting spatial information. The use of UAVs enables collection of a large amount of data on a relatively large area and enables faster mobilization to remote areas, emergency response, faster product generation and ability to collect data with multiple sensors. In addition, their use is cost effective and ideal for time-sensitive projects. Based on the collected data, it is possible to create different products such as ortho-photos, contours, topographic models, point clouds, 3D models, digital elevation and surface models and preform different analyses such as classification, slope and volumetric calculation, change detection, vegetation analysis, heat detections and others. The obtained products and analysis can be used for various purposes and improve projects in numerous fields such as agriculture, land survey, environmental studies, civil engineering, humanitarian efforts, real estate and other commercial domains. In order to successfully apply UAV technology for various projects, certain prerequisites are required. One of the main prerequisites is a properly designed flight plan. It is necessary to adequately ensure the correct collection of data during the flight depending on the purpose for which the flight is intended. Today, there are numerous applications for automatic flight planning, but each of them requires to correctly select certain parameters for successful product generation. Also, numerous data processing software have been developed. Most software contains built-in algorithms for obtaining products, and which software with the corresponding algorithm to choose depends on the planned purpose. The software differs from each other in processing algorithms, but also in the hardware requirements that computers must meet to use them. Although the use of UAVs is widely applicable today, there are also certain legislative restrictions for their use. In this presentation, several different applications for flight planning for different purposes will be presented, as well as several software for processing the collected data. A comparison of different application and software solutions will be given, with an assessment related to which software is best to use for a specific application. Also, a brief overview of the legal restrictions on UAV flying and examples of the obtained products for different purposes will be shown (download pdf).
Bio: Dr. Zvonimir Nevistić has a PhD in geodesy and geoinformatics from the Faculty of Geodesy, University of Zagreb, Croatia. In 2009, he enrolled in the bachelor study of Geodesy and Geoinformatics in Zagreb. He graduated in 2015 with the master thesis topic “Web map and tourist plan of the Bedekovčina municipality”. During his studies, he worked as a teaching demonstrator at the Department of Cartography and the Department of Spatial Information Management and was awarded with rector’s award for the project “Intensive field survey of the Kolan and surroundings”. After his studies, he was employed in the Agency for Payments in Agriculture, Fisheries and Rural Development in the remote sensing control sector, and then in the City of Zagreb in the property-legal relations sector of the land construction department.
Since 2016, he is employee at Faculty of Geodesy University of Zagreb as an assistant at the Department of Satellite Geodesy, and in the same year he enrolled postgraduate doctoral studies. While working at the Faculty, he was involved in the teaching of the following courses: Satellite positioning, Integrated systems in geomatics, Navigation, Engineering informatics, Geodetic astronomy and Space geodesy. In addition to teaching, he actively participates in numerous projects as a researcher technician and administrator. Projects in which he is actively involved are Erasmus + CBHE and SSA projects: BESTSDI, GEOBIZ, SEED4NA, UN4DDR and EO4GEO and Horizon 2020 TODO project. He is part of the Copernicus Academy & Relay team of the Faculty of Geodesy, and in 2019 he was one of the main organizers of the Copernicus Hackathon Zagreb. He participated in numerous domestic and international conferences, organized workshops, and meetings, published works in several journals, was the leader of 19 master theses and attended numerous international education and training courses. The topic of the doctoral thesis, made under the mentorship of prof. Ph.D. Željko Bačić is: Improving the availability and usability of space research data by modeling the spatial data infrastructure of celestial bodies. The main domains of research work are related to the application of GNSS and other modern satellite technologies, the use of drones and remote sensing in geodesy and geoinformatics, and SDI and smart cities. He actively speaks and writes English.
