Radio-based Multi-Robot Odometry and Relative Localization

TL;DR

This work addresses robust relative localization for a UGV-UAV team operating in GPS-denied environments by fusing inter-robot UWB range measurements with radar-based odometry in a 4-DOF pose-graph framework. It introduces three core components: a real-time 4-DOF nonlinear least-squares inter-robot transform estimator from anchor-tag UWB data, a radar odometry module with Doppler-velocity–driven ego-motion and GICP-based scan matching, and a multi-robot pose-graph that integrates inter-robot constraints, radar factors, and proprioceptive odometry within SITL and real-world validation, including a public Gazebo plugin and dataset. The results demonstrate that the inter-robot transform can be recovered without an initial guess, radar odometry provides robust velocity estimates and aids scan matching, and the fused system achieves globally consistent localization with reasonable errors (e.g., translation around 1 m and yaw a few degrees after convergence) at real-time rates. These findings imply that radio-based localization, augmented by factor-graph optimization, offers a practical path to robust multi-robot SLAM in challenging environments, with open-source resources enabling reproducibility and benchmarking.

Abstract

Radio-based methods such as Ultra-Wideband (UWB) and RAdio Detection And Ranging (radar), which have traditionally seen limited adoption in robotics, are experiencing a boost in popularity thanks to their robustness to harsh environmental conditions and cluttered environments. This work proposes a multi-robot UGV-UAV localization system that leverages the two technologies with inexpensive and readily-available sensors, such as Inertial Measurement Units (IMUs) and wheel encoders, to estimate the relative position of an aerial robot with respect to a ground robot. The first stage of the system pipeline includes a nonlinear optimization framework to trilaterate the location of the aerial platform based on UWB range data, and a radar pre-processing module with loosely coupled ego-motion estimation which has been adapted for a multi-robot scenario. Then, the pre-processed radar data as well as the relative transformation are fed to a pose-graph optimization framework with odometry and inter-robot constraints. The system, implemented for the Robotic Operating System (ROS 2) with the Ceres optimizer, has been validated in Software-in-the-Loop (SITL) simulations and in a real-world dataset. The proposed relative localization module outperforms state-of-the-art closed-form methods which are less robust to noise. Our SITL environment includes a custom Gazebo plugin for generating realistic UWB measurements modeled after real data. Conveniently, the proposed factor graph formulation makes the system readily extensible to full Simultaneous Localization And Mapping (SLAM). Finally, all the code and experimental data is publicly available to support reproducibility and to serve as a common open dataset for benchmarking.

Figures (9)

• Figure 1: Each robot describes a trajectory in its own local odometry frame while collecting range measurements (dashed lines) between two tags and four anchors. Measurement from different tags are represented in different colors.
• Figure 2: (a) and (c): Top and lateral view of the UGV used in the experiments. (b) and (d): Top and lateral view of the Matrice M210 used in the experiments. UWB sensors are highlighted in purple, and the radar sensors are highlighted in red.
• Figure 3: Measurements collected during the starting segment of the dataset trajectory for each pair, as compared with ground truth ranges.
• Figure 4: Overall architecture of the system.
• Figure 5: Comparison of the RMSE of the estimated relative transform in the three evaluated methods with $\sigma=2$ cm.