About us
Object Vision develops software and spatial models that support decision making in complex environments. Our work combines analytical rigour with technical depth: we build tools that governments, researchers and organisations use to understand spatial dynamics and evaluate future scenarios.
Our team brings together spatial economics, data analysis, modelling and software engineering. That mix is what lets us turn scientific insight into tools that survive contact with real policy questions. Reliability, transparency and methodological quality are not features we add at the end; they decide how we design a system in the first place.
Our main product is GeoDMS: an open-source framework for spatial data management, fast grid calculations and reproducible modelling workflows. All our software is open source, and we actively promote transparent modelling and the use of open data.
That position is not an idiosyncratic one. PBL sets out its own requirements for the models it uses: they must be scientifically sound, transparent about their assumptions and limitations, and open to review. We have built to that standard since long before it was written down, and it is steadily becoming the norm among the organisations we work for.
Thirty years of land-use modelling
1994, the Vrije Universiteit. The Faculty of Economics and Econometrics ran a research line on spatial data and GIS. Within it, together with Geodan, work began on SimEurope, a European land-use model. Maarten, then a PhD student, developed software libraries for network calculations, map algebra and accessibility modelling; Martin joined as an employee working in the same field. Those libraries became the foundation of what is now GeoDMS.
1996, adoption by the RIVM. The model was taken over by sector 5 of the RIVM as part of the LUMOS consortium (Land Use MOdelling System), where the national land-cover database LGN was combined with the land-use statistics BBG into one integrated map. That RIVM sector later became the Netherlands Environmental Assessment Agency ('Milieu- en Natuur Planbureau'; MNP), which in turn merged with the Spatial Planning Agency ('Ruimtelijk Planbureau'; RPB) in 2008 to form today's PBL ('Planbureau voor de Leefomgeving'). Our longest client relationship therefore predates the client's own name by more than a decade.
1997 and 1999, the Land Use Scanner. The model went into operation for national policy analysis, first described by Schotten et al. (1997). Two years later Hilferink and Rietveld (1999) set out its formal foundation and raised the resolution from 500 to 100 metres. That paper introduced the logit approach to spatial allocation, borrowing discrete choice theory from transport modelling: where a transport model chooses a route, this model chooses a type of land use for a location.
Early 2000s, the joint venture in Haarlem. A joint venture with the software company YUSE GSO, based in Haarlem, allowed the software to be extended with a larger team of experienced developers, and broadened the range of applications built on it.
2006, Object Vision B.V. We returned to the Vrije Universiteit Amsterdam, where our office still is, and formally established the company. The focus shifted from core software development towards applied projects: land-use modelling, accessibility analysis, environmental assessment and data pipelines. GeoDMS kept evolving, with more vector operations, better performance and integration with web-based applications.
Europe and beyond. The EU-ClueScanner, built for DG Environment and the Joint Research Centre of the European Commission, became the predecessor of LUISA, the Commission's territorial modelling platform for ex-ante assessment of policies with a spatial impact. The Bangladesh Delta Scanner has been used by Deltares and CEGIS in Dhaka since 2014 for land-use and water management scenarios, and a Land Use Scanner for Germany was developed with the Federal Office for Building and Regional Planning ('Bundesamt für Bauwesen und Raumordnung'; BBR).
2007 onwards, an audit and an economic turn. An international scientific audit (Timmermans et al., 2007) recommended tying the model logic more firmly to economic and policy logic. That led to an economic basis for suitability, where bid prices for land drive local suitability and financial instruments such as taxes and subsidies can be modelled explicitly.
Growth. In 2018 Jip joined, combining consultancy with a part-time PhD in spatial economics at the VU. In 2021 Erik joined, strengthening both software development and modelling capacity.
2021 onwards, RuimteScanner 2.0. The single land-use layer was split into three: actors, objects and land use. The model became open source as RSopen, and was first applied in the Planmonitor NOVI 2023 and the Ruimtelijke Verkenning 2023. It is now also used extensively by Deltares in their spatial modelling work.
Governance
In 2025 we changed how the company is governed. Object Vision B.V. continues to exist and to do the work it always did, but it is now wholly owned by Coöperatieve Vereniging Object Vision U.A. Control therefore sits with the cooperative rather than with individual shareholders, and its certificates are held by the professionals working in the company.
The practical effect is continuity. Decisions about direction are taken by the people who design, build and maintain the models, which keeps the company independent and its horizon long. For clients and partners nothing changes in what we do or how we work.