Redundant and Loosely Coupled LiDAR-Wi-Fi Integration for Robust Global Localization in Autonomous Mobile Robotics

TL;DR

This paper presents a framework addressing the challenge of global localization in autonomous mobile robotics by integrating LiDAR-based descriptors and Wi-Fi finger-printing in a pre-mapped environment with a redundancy-based approach that enhances the system's overall robustness and accuracy.

Abstract

This paper presents a framework addressing the challenge of global localization in autonomous mobile robotics by integrating LiDAR-based descriptors and Wi-Fi fingerprinting in a pre-mapped environment. This is motivated by the increasing demand for reliable localization in complex scenarios, such as urban areas or underground mines, requiring robust systems able to overcome limitations faced by traditional Global Navigation Satellite System (GNSS)-based localization methods. By leveraging the complementary strengths of LiDAR and Wi-Fi sensors used to generate predictions and evaluate the confidence of each prediction as an indicator of potential degradation, we propose a redundancy-based approach that enhances the system's overall robustness and accuracy. The proposed framework allows independent operation of the LiDAR and Wi-Fi sensors, ensuring system redundancy. By combining the predictions while considering their confidence levels, we achieve enhanced and consistent performance in localization tasks.

Figures (5)

• Figure 1: Point cloud map generated from an experiment showcased in this article, demonstrating the queried point cloud, and showcasing several mobile robots equipped with the novel and lightweight proposed solution.
• Figure 2: An overview of the proposed system architecture, divided into four main components. (A) LiDAR-based place recognition; (B) Wi-Fi fingerprinting; (C) best candidate selection; and (D) ICP Point Cloud Registration.
• Figure 3: Example of correlation matrices $\mathcal{C}_{i,j}$, where the empty cells correspond to the case $A_{t,i} \neq A_{k,j}$, while the colored cells correspond to the case $A_{t,i} = A_{k,j}$.
• Figure 4: Top plots showcase the results from the real-world experiments in terms of place recognition for an increasing number $N$ of nearest place candidates retrieved from the map. A candidate is considered correct if it is within a $3$ m radius from the corresponding pose in the database. Bottom plots demonstrate the mean distance of the predicted pose to the ground truth from the map, for each candidate, with the addition of the min and max deviations.
• Figure 5: Trajectory for each experiment, based on the top-1 candidate (green points represent the correct predictions, red points denote the incorrect ones).