RCA: Ride Comfort-Aware Visual Navigation via Self-Supervised Learning

TL;DR

This paper introduces a self-supervised learning framework that models ride comfort explicitly from visual and proprioceptive data, improving navigation safety and comfort for wheelchair users in shared autonomy settings.

Contribution

It presents a novel approach to predict ride comfort directly from first-person images using self-supervised learning, enhancing traversability analysis for safer, more comfortable navigation.

Findings

Navigation system achieves human-preferred ride comfort

Proposed model accurately predicts perceived ride difficulty

System effectively integrates terrain appearance and vehicle states

Abstract

Under shared autonomy, wheelchair users expect vehicles to provide safe and comfortable rides while following users high-level navigation plans. To find such a path, vehicles negotiate with different terrains and assess their traversal difficulty. Most prior works model surroundings either through geometric representations or semantic classifications, which do not reflect perceived motion intensity and ride comfort in downstream navigation tasks. We propose to model ride comfort explicitly in traversability analysis using proprioceptive sensing. We develop a self-supervised learning framework to predict traversability costmap from first-person-view images by leveraging vehicle states as training signals. Our approach estimates how the vehicle would feel if traversing over based on terrain appearances. We then show our navigation system provides human-preferred ride comfort through robot experiments together with a human evaluation study.