Adaptive Robot Localization with Ultra-wideband Novelty Detection

2025-05-09

AI Summary

In indoor cluttered environments, UWB-based robot localization suffers from degraded accuracy due to multipath propagation, signal reflections, and non-line-of-sight (NLOS) interference. To address this, we propose a robust adaptive localization framework that integrates a semi-supervised autoencoder for UWB ranging anomaly detection and novelty identification, enabling dynamic adjustment of the observation covariance and bias estimation in an extended Kalman filter (EKF). Our approach is the first to deeply fuse UWB data credibility modeling with adaptive EKF across both spatial and temporal dimensions, achieving generalization to unseen NLOS scenarios without retraining. Experimental validation on a real robotic platform demonstrates that, under NLOS conditions, the proposed method reduces average localization error by nearly 60% and decreases absolute error by over 25 cm, significantly enhancing localization robustness and generalizability in complex indoor settings.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Machine Learning: Adversarial Learning & Robustness
• Natural Language Processing: Safety and Robustness

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Location- and context-aware Web and WoT applications and services
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Ultra-wideband (UWB) technology has shown remarkable potential as a low-cost general solution for robot localization. However, limitations of the UWB signal for precise positioning arise from the disturbances caused by the environment itself, due to reflectance, multi-path effect, and Non-Line-of-Sight (NLOS) conditions. This problem is emphasized in cluttered indoor spaces where service robotic platforms usually operate. Both model-based and learning-based methods are currently under investigation to precisely predict the UWB error patterns. Despite the great capability in approximating strong non-linearity, learning-based methods often do not consider environmental factors and require data collection and re-training for unseen data distributions, making them not practically feasible on a large scale. The goal of this research is to develop a robust and adaptive UWB localization method for indoor confined spaces. A novelty detection technique is used to recognize outlier conditions from nominal UWB range data with a semi-supervised autoencoder. Then, the obtained novelty scores are combined with an Extended Kalman filter, leveraging a dynamic estimation of covariance and bias error for each range measurement received from the UWB anchors. The resulting solution is a compact, flexible, and robust system which enables the localization system to adapt the trustworthiness of UWB data spatially and temporally in the environment. The extensive experimentation conducted with a real robot in a wide range of testing scenarios demonstrates the advantages and benefits of the proposed solution in indoor cluttered spaces presenting NLoS conditions, reaching an average improvement of almost 60% and greater than 25cm of absolute positioning error.

Problem

• Address UWB signal disturbances in cluttered indoor environments
• Develop adaptive localization without extensive retraining for new data
• Improve robot positioning accuracy in NLOS conditions

Innovation

• UWB novelty detection with semi-supervised autoencoder
• Dynamic covariance and bias error estimation
• Extended Kalman filter integration for robust localization