Kadastrale Kaart Next — the Netherlands Rebuilds Its National Cadastral Map With AI, Under Independent Scrutiny

Kadaster, the Dutch national land registry and mapping agency, is digitising roughly 200 years of handwritten field survey sketches ('veldwerken') to rebuild the country's cadastral boundary map, correcting scale-conversion shifts and connection errors accumulated across historical paper originals. The Kadastrale Kaart Next (KKN) programme applies AI, including automated interpretation of handwritten measurement data, to reprocess around 5 million field sketches with two contracted supplier partners, on a budget of €35 million, and the system is formally listed on the Dutch government's public algorithm register.

Before rollout, the programme was independently reviewed by the Dutch government's ICT assessment board (Adviescollege ICT-toetsing / BIT), which concluded the programme was carefully prepared technically but flagged that it remained uncertain when and to what extent the programme's benefits would be realised, criticised the cost-benefit analysis as insufficiently concrete, and warned of delay risk from a lack of integrated management. Kadaster itself acknowledges that small inaccuracies can still arise from the historical maps' scale and alignment issues, and the workflow keeps a human in the loop rather than fully automating map reconstruction.

According to the contractor's own technical account, AI-assisted processing cut handling time per sketch from roughly 90 minutes to about 18 minutes. Full production was targeted for the end of 2026 as of the programme's own reporting.

As of the most recent public reporting, KKN had not yet reached full production, making this a case of a transparently disclosed, independently audited AI programme whose real-world benefits are still being tested rather than proven.