Multi-Agent Collaborative GNSS for Intelligent Transportation Systems

This program develops reliable multi-agent collaborative GNSS positioning for intelligent transportation systems — vehicles in an urban convoy share their raw GNSS measurements through inter-vehicle ranging, so that collectively they recover position fixes that individual vehicles couldn’t trust on their own. The work spans the full stack: measurement-quality assessment in urban canyons, environment-aware NLOS handling, factor-graph optimization to fuse heterogeneous measurements, and integrity monitoring to bound the trust of the shared solution.

Why it matters. Single-vehicle GNSS in dense cities is unreliable. Multi-agent collaboration is a route to both accuracy and integrity at once — if peer vehicles can detect each other’s faults, the swarm becomes more trustworthy than any single sensor.

Funding & recognition.

• RGC Research Impact Fund R5009-21 — HK$4.43M total project cost (RGC contribution HK$3.1M + 30% institutional matching), Project Coordinator (the only RIF held as PC in Li-Ta’s career to date), 2022–2025
• 2023 ION Per Enge Early Achievement Award — second-ever Asian-university recipient, citing the foundational urban-GNSS-reliability + integrity work
• 2024 Most-Cited Paper in NAVIGATION — Factor Graph Optimization for GNSS/INS Integration: A Comparison with the Extended Kalman Filter (Wen, Pfeifer, Bai, Hsu) — the FGO foundation that the multi-agent collaboration builds on
• PolyU Presidential Award in Knowledge Transfer 2022–24 — recognizing the broader KT impact of the multi-agent collaborative GNSS research line and its industry translations
• US12123961B2 patent (granted 2024) — 3D LiDAR Aided GNSS NLOS Detection and Correction

Outputs (representative).

• UrbanNav — open multisensory benchmark dataset for multi-agent positioning research
• Integrity-Constrained Factor Graph Optimization for GNSS (Xia, Wen, Hsu, NAVIGATION 2024)
• Principal Gaussian Overbound for Heavy-tailed Pseudorange Error Bounding (Yan, Zhong, Hsu, IEEE TAES 2024)
• 3D Mapping Database Aided GNSS-based Collaborative Positioning via FGO (Zhang, Ng, Hsu, IEEE TITS 2020)
• GNSS RUMS — Realistic Urban Multi-Agent Simulator for Collaborative Positioning Research (Zhang et al., Remote Sensing 2021)
• Multiple Faults Isolation for Multi-Constellation GNSS Positioning (Yan et al., IEEE Sensors J 2025)