How IMU Drift Influences Multi-Radar Inertial Odometry for Ground Robots in Subterranean Terrains

Abstract

Reliable radar inertial odometry (RIO) requires mitigating IMU bias drift, a challenge that intensifies in subterranean environments due to extreme temperatures and gravity-induced accelerations. Cost-effective IMUs such as the Pixhawk, when paired with FMCW TI IWR6843AOP EVM radars, suffer from drift-induced degradation compounded by sparse, noisy, and flickering radar returns, making fusion less stable than LiDAR-based odometry. Yet, LiDAR fails under smoke, dust, and aerosols, whereas FMCW radars remain compact, lightweight, cost-effective, and robust in these situations. To address these challenges, we propose a two-stage MRIO framework that combines an IMU bias estimator for resilient localization and mapping in GPS-denied subterranean environments affected by smoke. Radar-based ego-velocity estimation is formulated through a least-squares approach and incorporated into an EKF for online IMU bias correction; the corrected IMU accelerations are fused with heterogeneous measurements from multiple radars and an IMU to refine odometry. The proposed framework further supports radar-only mapping by exploiting the robot's estimated translational and rotational displacements. In subterranean field trials, MRIO delivers robust localization and mapping, outperforming EKF-RIO. It maintains accuracy across cost-efficient FMCW radar setups and different IMUs, showing resilience with Pixhawk and higher-grade units such as VectorNav. The implementation will be provided as an open-source resource to the community (code available at https://github.com/LTU-RAI/MRIO

Figures (7)

• Figure 1: Subterranean test environment with ground robot equipped with FMCW TI IWR6843AOP EVM radars and two IMU configurations. Experiments in a sloped, extreme-cold tunnel show the proposed MRIO framework outperforming state-of-the-art EKF-RIO with online calibration across all IMU setups.
• Figure 2: A schematic of the Multi-Radar Inertial Odometry (MRIO) framework, which integrates inputs from an IMU and multiple radars mounted on a mobile robotic platform to obtain odometry, ego velocity, and a radar point-cloud map. Radar scans, containing target points and Doppler velocities, are processed via a least-squares filter to obtain ego velocity, which is then refined using a two-stage filtering approach for accurate localization and mapping.
• Figure 3: On the left: The experimental setup includes six mmWave TI-IWR6843AoP compact radars (labeled $R_1$ to $R_6$), arranged in an ensemble configuration with LiDAR and IMU, mounted on the Pioneer 3-AT2 rover platform. Here, radars $R_1-R_4$ are mounted on the same plane, while radars $R_5-R_6$ are positioned at a slanted angle, capturing the perception of the ceiling from both the front and rear sides. On the right: Visualization of radar point clouds captured at a single moment using the proposed multi-radar setup. The target points are colour-coded, with each colour representing different radar target points${}^{\mathcal{R}_i}_{}\bm{P}_{p_t} \forall i= 1\dots 6$
• Figure 4: Localization performance comparison across IMUs using MRIO and EKF– RIO.
• Figure 5: Experiments in the SubT tunnel with PX4 and VectorNav IMUs demonstrate that the proposed framework achieves robust localization and mapping performance compared to LIO-SLAM across all scenarios.