A Cost-Effective Radar-Visual-Inertial SLAM System for Indoor Land Vehicle Navigation

Emma Dawson, Paulo Ricardo Marques de Araujo, Sidney Givigi, Aboelmagd Noureldin

Peer Reviewed

Abstract:	Due to the increasing demand for safety, comfort, and convenience for both drivers and passengers, modern vehicles are equipped with a diverse set of advanced driver assistance systems (ADAS), in which automotive radars have been successfully applied to tasks such as adaptive cruise control, collision avoidance, and blind spot detection. Advances in radar technology have motivated research into radars applied to other automotive tasks, such as positioning. There exists extensive development and open source support for SLAM systems for LiDAR point clouds. However, point clouds generated from current automotive radars are insufficiently dense for use in SLAM pipelines developed for LiDAR data. This work presents three radar scan densification algorithms, aimed at enhancing the suitability of automotive radar for use in existing SLAM systems. The effectiveness of these densification methods is assessed using the established LVI-SAM pipeline, which integrates radar point clouds, an inertial measurement unit, and monocular cameras. Different parameter configurations are tested for each densification method. The pipeline is tested using a land vehicle setup inside covered parking garages. The root mean squared positioning error achieved by the best-performing densification method was 0.54 m in an indoor parking garage, which is a challenging environment for automotive radars due to the presence of multi-path effects and ghost detections. The results indicate the potential of radar systems for SLAM pipelines, with the additional benefits of the low cost and wide availability of radar in vehicles.
Published in:	Proceedings of the 38th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2025) September 8 - 12, 2025 Hilton Baltimore Inner Harbor Baltimore, Maryland
Pages:	1673 - 1686
Cite this article:	Dawson, Emma, de Araujo, Paulo Ricardo Marques, Givigi, Sidney, Noureldin, Aboelmagd, "A Cost-Effective Radar-Visual-Inertial SLAM System for Indoor Land Vehicle Navigation," Proceedings of the 38th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2025), Baltimore, Maryland, September 2025, pp. 1673-1686. https://doi.org/10.33012/2025.20279
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In