Readar Aerial Imagery Data Extraction

Readar is a Dutch geospatial data company (Utrecht, Netherlands; founded 2016) that derives nationwide feature datasets from aerial photographs using remote sensing and machine learning. The core service ingests licensed aerial imagery covering the Netherlands and applies automated or semi-automatic ML pipelines to extract structured geospatial information at address level.

Output datasets include point clouds from stereo aerial photogrammetry, TrueOrtho mosaics, 3D building and roof models, building volumes and floor areas, solar panel presence and potential, suspected asbestos roof identification, and change detection datasets. All outputs are linkable to property addresses and delivered via real-time API or batch file export. Two processing modes are available: a fully automatic pipeline and a semi-automatic workflow in which ML algorithms identify candidate objects for human tagger review, with sample size set by statistical confidence requirements.

Readar also processes client-supplied aerial imagery and operates in Belgium. The company was incubated at ESA BIC Noordwijk (2017-2018). Customers include municipalities, property insurers, energy companies, banks, grid operators, and installation businesses.^[1][2][3][4]