LT-Exosense: A Vision-centric Multi-session Mapping System for Lifelong Safe Navigation of Exoskeletons

2025-10-25

AI Summary

To address the challenge of perception reliability for long-term autonomous navigation of lower-limb exoskeletons in dynamic environments, this paper proposes a vision-centric multi-session mapping system. The method introduces cross-session SLAM to exoskeleton navigation for the first time, enabling incremental spatial knowledge fusion, environment change detection, and online persistent map updating—thereby constructing a globally consistent, dynamically evolving semantic map that supports obstacle-aware path planning and reversible recovery of historical trajectories. By tightly fusing monocular camera and IMU data, the system achieves high-precision localization (average point-to-point error < 5 cm) and lightweight real-time mapping. Experimental validation in realistic indoor environments demonstrates robust multi-session map scalability and responsive adaptation to environmental dynamics. The proposed framework significantly enhances the safety and autonomy of exoskeletons during extended operational periods.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Location- and context-aware Web and WoT applications and services
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Self-balancing exoskeletons offer a promising mobility solution for individuals with lower-limb disabilities. For reliable long-term operation, these exoskeletons require a perception system that is effective in changing environments. In this work, we introduce LT-Exosense, a vision-centric, multi-session mapping system designed to support long-term (semi)-autonomous navigation for exoskeleton users. LT-Exosense extends single-session mapping capabilities by incrementally fusing spatial knowledge across multiple sessions, detecting environmental changes, and updating a persistent global map. This representation enables intelligent path planning, which can adapt to newly observed obstacles and can recover previous routes when obstructions are removed. We validate LT-Exosense through several real-world experiments, demonstrating a scalable multi-session map that achieves an average point-to-point error below 5 cm when compared to ground-truth laser scans. We also illustrate the potential application of adaptive path planning in dynamically changing indoor environments.

Problem

• Developing lifelong navigation for exoskeletons in changing environments
• Creating multi-session mapping that fuses spatial knowledge incrementally
• Enabling adaptive path planning around obstacles in dynamic settings

Innovation

• Vision-centric multi-session mapping system
• Incrementally fuses spatial knowledge across sessions
• Adaptive path planning for dynamic environments