Videos
UrbanNav is an open multisensory benchmark for testing navigation algorithms in dense urban environments. It provides raw GNSS, IMU, LiDAR, camera, and ground-truth data from challenging Asian urban canyon scenarios — supporting research in GNSS positioning, sensor fusion, SLAM, and integrity monitoring.
Why it matters. Real urban navigation fails because GNSS, perception, and low-cost sensors fail together. UrbanNav gives the community a shared benchmark for those joint failures, and has been adopted as a reference dataset across the positioning and navigation community.
Outputs: open dataset, GitHub repository, benchmark publications (NAVIGATION 2023; ION GNSS+ 2021), tutorials, and integration with the GraphGNSSLib factor-graph platform.
Funding: Google, RGC, and other sources
Started: 2019