Safety-certifiable Multi-Sensor Fusion

Our Approach

1. Robust fusion in dynamic scenes

   Detect and remove moving objects and outliers so that LiDAR, visual and inertial odometry stay consistent in crowded streets.

2. Safety-quantifiable localisation

   Line- and plane-feature localisation against 3D prior maps, with a quantified error bound for every pose.

3. Integrity monitoring for multi-sensor systems

   Fault modes, alert limits and protection levels adapted from aviation ARAIM to urban vehicles, aircraft and wearables.

Research in Action

Systems we have built and tested with partners. Open a card for the problem, our approach, results and photos.

Safety-Certified Multi-Source Fusion Positioning for Autonomous Vehicles

Integrity monitoring brought from aviation into urban autonomous-vehicle positioning.

• ARAIM integrity method

Problem, approach & results

ProblemAV positioning in cities fails silently: sensors degrade together and no alert limit tells the planner when to stop trusting it.

Approach

• Urban alert limits derived from lane geometry and vehicle size
• Fault modes catalogued for GNSS, LiDAR and camera in urban scenes
• ARAIM-style fault detection, exclusion and protection levels for multi-sensor fusion

ResultsOperational and safety-requirement report, system design and final test report.

Autonomous Take-off/Landing and Obstacle Avoidance for Civil Aircraft

Multi-source fusion for safety-critical take-off, landing and obstacle avoidance of civil aircraft.

• ±0.3 m runway-centreline target
• 10 test scenarios
• TRL 4→5 technology-readiness target

Problem, approach & results

ProblemInstrument landing depends on ground signals, single sensors can be jammed or spoofed, and vision degrades at night, in haze and in glare.

Approach

• Multispectral vision (visible, infrared, event camera) fused with IMU and GNSS
• Visual-inertial navigation with RAIM-style integrity checks and robust MPC control
• Digital-twin test bench with 10 scenarios: night, haze, GPS failure, wind gusts and joint faults

ResultsMultispectral perception and VIO prototype, robust MPC module, digital-twin platform and a TRL-5 validation report.

AI-Assisted Inertial Navigation System

Deep-learning inertial odometry for phones and wearables, now deployed on consumer devices.

• 60K+ devices deployed
• Certified technology application

Problem, approach & results

ProblemPhones and wearables must keep tracking users where GNSS is weak, using only noisy IMU and magnetometer data.

Approach

• Deep-learning inertial odometry from accelerometer, gyroscope and magnetometer
• Trajectory output with confidence for downstream fusion
• Adaptive error-state Kalman filter for wearable attitude

ResultsCertificate of technology application from the partner.

Maximum Consensus Integration of GNSS and LiDAR for Urban Navigation

Outlier-robust urban navigation with Leibniz University Hannover: keep only the GNSS and LiDAR measurements that agree with each other.

• 2 universities
• Exchange visits both ways
• Robust outlier rejection

Problem, approach & results

ProblemGNSS and LiDAR measurements in cities contain many outliers that break least-squares and Kalman estimators.

Approach

• Maximum-consensus estimation that keeps only mutually consistent measurements
• Joint GNSS–LiDAR formulation for urban vehicles
• Student and staff exchange between Hong Kong and Hannover

ResultsBuilt a lasting research link with Leibniz University Hannover.

Video Demonstrations

Selected Publications

2025

• POPL-SLAM: A Pose-Only Representation-Based Visual-Inertial SLAM With Point and Structural Line Features.
  Li, T., Han, B., Yan, D., Wen, W., Wang, Z., Shi, C.
  IEEE Transactions on Aerospace and Electronic Systems, 62, 1509-1525, 2025. (IF: 5.7, JCR Q1)
• Safety-quantifiable Line Feature-based Monocular Visual Localization with 3D Prior Map.
  Zheng, X., Wen, W.*, Hsu, L.T.
  IEEE Transactions on Intelligent Transportation Systems, 2025. (IF: 8.4, JCR Q1, Citations: 3)
• Fault Detection Algorithm for Gaussian Mixture Noises: An Application in Lidar/IMU Integrated Localization Systems.
  Yan, P., Li, Z., Huang, F., Wen, W., Hsu, L.T.
  NAVIGATION: Journal of the Institute of Navigation, 72(1), 2025. (IF: 3.1, JCR Q1, Citations: 6)
• Multi-Sensor Plug-and-Play Navigation Based on Resilient Information Filter.
  Meng, Q., Su, C., Jiang, Y., Wen, W., Meng, X.
  IEEE Sensors Journal, 2025. (IF: 4.5, JCR Q1)
• A Novel Lie Group-based Reliable IMM Estimation Method for SINS/GNSS/OD/NHC Integrated Navigation in Complex Environments.
  Du, S., Huang, Y., Wen, W., Zhang, Y.
  IEEE Internet of Things Journal, 2025. (IF: 8.9, JCR Q1, Citations: 5)
• Graph-Based Indoor 3D Pedestrian Location Tracking With Inertial-Only Perception.
  Bai, S., Wen, W.*, Su, D., Hsu, L.T.
  IEEE Transactions on Mobile Computing, 2025. (IF: 9.2, JCR Q1, Citations: 5)

2024

• Safety-Quantifiable Planar-Feature-based LiDAR Localization with a Prior Map for Intelligent Vehicles in Urban Scenarios.
  Zhang, J., Liu, X., Wen, W.*, Hsu, L.T.
  IEEE Transactions on Intelligent Vehicles, 2024. (IF: 14.3, JCR Q1, Citations: 2)
• Factor Graph Optimization-Based Smartphone IMU-Only Indoor SLAM With Multi-Hypothesis Turning Behavior Loop Closures.
  Bai, S., Wen, W.*, Hsu, L.T., Yang, P.
  IEEE Transactions on Aerospace and Electronic Systems, 2024. (IF: 5.7, JCR Q1, Citations: 9)
• Tightly-coupled Visual/Inertial/Map Integration with Observability Analysis for Reliable Localization of Intelligent Vehicles.
  Zheng, X., Wen, W.*, Hsu, L.T.
  IEEE Transactions on Intelligent Vehicles, 2024. (IF: 14.3, JCR Q1, Citations: 3)
• Integrity-Constrained Factor Graph Optimization for GNSS Positioning in Urban Canyons.
  Xia, X., Wen, W., Hsu, L.T.*
  NAVIGATION: Journal of the Institute of Navigation, 2024. (IF: 3.1, JCR Q1, Citations: 5)
• FGO-MFI: Factor Graph Optimization-based Multi-sensor Fusion and Integration for Reliable Localization.
  Zhu, J., Zhuo, G., Xia, X.*, Wen, W., Xiong, L., Leng, B., Liu, W.
  Measurement Science and Technology, 35(8), 086303, 2024. (IF: 3.4, JCR Q1, Citations: 7)

2023

• GLIO: Tightly-coupled GNSS/LiDAR/IMU Integration for Continuous and Drift-free State Estimation of Intelligent Vehicles in Urban Areas.
  Liu, X., Wen, W.*, Hsu, L.T.
  IEEE Transactions on Intelligent Vehicles, 2023. (IF: 14.3, JCR Q1, Citations: 52)
• Low-cost Solid-state LiDAR/Inertial Based Localization with Prior Map for Autonomous Systems in Urban Scenarios.
  Zhong, Y., Huang, F., Zhang, J., Wen, W.*, Hsu, L.T.
  IET Intelligent Transport Systems, 17(3), 474-486, 2023. (IF: 2.3, JCR Q2, Citations: 9)

2018–2022

• AGPC-SLAM: Absolute Ground Plane Constrained 3D Lidar SLAM.
  Wen, W., & Hsu, L.T.
  NAVIGATION: Journal of the Institute of Navigation, 69(3), 2022.
• Coarse-to-Fine Loosely-Coupled LiDAR-Inertial Odometry for Urban Positioning and Mapping.
  Zhang, J., Wen, W.*, Huang, F., Chen, X., Hsu, L.T.
  Remote Sensing, 13, 2371, 2021.
• Point Wise or Feature Wise? A Benchmark Comparison of Publicly Available Lidar Odometry Algorithms in Urban Canyons.
  Huang, F., Wen, W., Zhang, J., Hsu, L.T.*
  IEEE Intelligent Transportation Systems Magazine, 2021.
• Robust Visual-Inertial Integrated Navigation System Aided by Online Sensor Model Adaption for Autonomous Ground Vehicles in Urban Areas.
  Bai, X., Wen, W., Hsu, L.T.
  Remote Sensing, 12(10), 1686, 2020.
• Performance Analysis of NDT-based Graph SLAM for Autonomous Vehicle in Diverse Typical Driving Scenarios of Hong Kong.
  Wen, W., Hsu, L.T., Zhang, G.
  Sensors, 18, 3928, 2018.

Full publication list →