Delft University of Technology (TU Delft) brings advanced expertise in 3D geoinformation, urban modelling and machine learning to MultiRoofs. Its 3D Geoinformation Group develops open-data and open-source solutions for modelling cities, buildings and landscapes in 3D. Building on projects such as 3D BAG, which models all 10 million buildings in the Netherlands in 3D, the team combines cutting-edge research with extensive experience in developing practical tools and international standards for urban geospatial data.

Within MultiRoofs, TU Delft plays a key role in the development of the Digital Twin, adapting its 3D building reconstruction methodology so that it can work with the diverse datasets available across different cities and countries in North-West Europe. The team also develops machine learning methods to classify rooftops and identify characteristics such as their shape and surface materials, helping cities better understand their rooftop potential and identify suitable uses. In addition, TU Delft works to make the resulting data interoperable, reusable and scalable by extending CityJSON, the international standard for 3D city models, to incorporate the specific requirements of MultiRoofs.