Work

Intressenters uppfattningar om digitalisering och automatisering i geospatiala planering och dataproduktionsarbetsflöden

Year: 2026
Type: dissertation
Abstract: Originally published: https://urn.fi/URN:ISBN:978-952-64-3335-6 The successful adoption of information systems depends on stakeholders’ perceptions of their usefulness, reliability, and understandability. These perceptions are particularly important when technologies are introduced to workflows that transition from manual or analogue methods to digital or automated systems. The research presented in this thesis examined stakeholders’ perceptions of two emerging geospatial technologies operating in different contexts: 1) collaborative geographic information systems (CGIS) used in maritime spatial planning (MSP) workshops, and 2) artificial intelligence (AI) used in the production of data for topographic databases (TDBs). The research examined system characteristics that influence MSP stakeholders’ perceptions of the utility, ease of use, and understandability of CGIS in MSP workshops and Finnish TDB users’ perceptions of trust, risks, transparency, and understandability of AI used in geospatial data production. Furthermore, the research identified challenges and limitations that may hinder successful adoption of the two technologies in the given context. The research used mixed methods, including prototype implementation, workshop observations, questionnaires, and analysis of AI-generated spatial outputs. In three of the four studies, research data were collected from real-world MSP stakeholders or TDB users. Understanding stakeholder perceptions is crucial for designing geospatial technologies that are both technically effective and socially acceptable. The results show that CGIS tools that support exploration, visual analysis, and creation of geospatial data are perceived as useful for various tasks in collaborative MSP settings. They also suggest that automated spatial analysis adds complexity and introduces issues with comprehension and ease of use. The integration of spatial analysis tools into CGIS can mitigate some challenges with using spatial analysis tools in collaborative settings by increasing methodological flexibility and offering additional support tools. Commonly reported machine learning metrics do not adequately explain spatial data quality metrics, which challenges the interpretation of data quality in AI-based TDB data production. Stakeholders generally perceive AI-based automation as trustworthy in producing high-quality data for TDBs if validating outcomes is done manually. However, stakeholders have many concerns about the security, privacy, and potential misuse of AI models. The results further indicate that a lack of transparency and comprehension can be barriers to the acceptance of systems in geospatial data production workflows. Overall, the findings in this thesis highlight the key factors shaping stakeholders’ trust, perceived utility, and acceptance of CGIS and AI technologies that are used in the digitalisation and automation of geospatial workflows. The research also identifies challenges that can be barriers to the technologies’ adoption. The thesis contributes new knowledge by addressing a gap in research on the adoption of geospatial technologies, particularly regarding stakeholder perceptions of CGIS and AI-based geospatial data production. Osajulkaisut [Publication 1]: Koski, C., Rönneberg, M., Kettunen, P., Eliasen, S., Hansen, H. S., & Oksanen, J. (2021). Utility of collaborative GIS for maritime spatial planning: Design and evaluation of Baltic Explorer. Transactions in GIS, 25(3), 1347-1365. https://doi.org/10.1111/tgis.12732 [Publication 2]: Koski, C., Rönneberg, M., Kettunen, P., Armoškaitė, A., Strake, S., & Oksanen, J. (2021). User experiences of using a spatial analysis tool in collaborative GIS for maritime spatial planning. Transactions in GIS, 25(4), 1809-1824. https://doi.org/10.1111/tgis.12827 [Publication 3]: Koski, C., Kettunen, P., Poutanen, J., Zhu, L., & Oksanen, J. (2023). Mapping small watercourses from DEMs with deep learning—exploring the causes of false predictions. Remote Sensing, 15(11), 2776. https://doi.org/10.3390/rs15112776 [Publication 4]: Koski, C., Kettunen, P., Tenkanen, H., & Oksanen, J., “Trust in artificial intelligence to extract geospatial features for topographic databases”, 35 pages, in press (August 2026), International Journal of Geographical Information Science. https://doi.org/10.1080/13658816.2026.2716279
Institution: Aalto University
Language: English
Open Access status: green
OpenAlex ID: W7213486929
Cites: 151
Cited by:
Related to: 9