Using AI-powered techniques to generate high-definition road transport network data from aerial imagery

Important results from the project

The project achieved its core objectives by further developing and validating Tapestry for Swedish and Scandinavian conditions. The target of at least 90% accuracy for motor vehicle networks was reached, and a first methodology for pedestrian and cycling infrastructure was established. The project also delivered a functional API, stronger user alignment, and laid the foundation for a new full-scale FFI project.

Expected long term effects

The project contributes to faster, more cost-efficient, and scalable generation of high-resolution transport network data. This improves the foundation for more accurate analysis, planning, and decision-making in the transport system. By enabling multimodal analysis and more up-to-date and quality-assured network data, it also supports the development of a more efficient, sustainable, inclusive, accessible, and resilient transport system.

Approach and implementation

The project was carried out as a pre-study with three main components: project management, technical development, and stakeholder engagement. The structure worked well and the project was implemented according to plan and within the timeframe. The technical objectives were achieved, resulting in concrete and validated outcomes. The dialogue with users and partners helped strengthen the project’s relevance, ensure alignment with real-world needs, and clarify the direction for further development.