A Multi-Layer Sim-to-Real Framework for Gaze-Driven Assistive Neck Exoskeletons
Colin Rubow, Eric Brewer, Ian Bales, Haohan Zhang, Daniel Brown
AI summary
Problem
Controlling assistive neck exoskeletons for Dropped Head Syndrome patients is difficult due to the challenge of predicting intended head movement from eye gaze and the safety risks of testing controllers directly on humans.
Approach
The authors propose a multi-layer evaluation pipeline that tests gaze-driven control models across decreasing levels of abstraction—simulation, virtual reality, and physical hardware—to safely and efficiently identify the best controllers before human deployment.
Key results
- A multi-layer controller selection framework that safely filters poor models early in development
- Successful rejection of four unstable controllers via autoregressive simulation before VR testing
- Two novel gaze-driven models achieve strong performance on physical hardware
- User studies show no single controller is universally preferred, emphasizing the need for personalization
Why it matters
This work accelerates the safe development of intuitive assistive robots for neurological patients by providing a reliable evaluation pipeline and demonstrating that personalized gaze-driven control is essential.
Abstract
Dropped head syndrome, caused by neck muscle weakness from neurological diseases, severely impairs an indi- vidual’s ability to support and move their head, causing pain and making everyday tasks challenging. Our long-term goal is to develop an assistive powered neck exoskeleton that restores natural movement. However, predicting a user’s intended head movement remains a key challenge. We leverage virtual reality (VR) to collect coupled eye and head movement data from healthy individuals to train models capable of predicting head movement based solely on eye gaze. We also propose a novel multi-layer controller selection framework, where head control strategies are evaluated across decreasing levels of abstrac- tion—from simulation and VR to a physical neck exoskeleton. This pipeline effectively rejects poor-performing controllers early, identifying two novel gaze-driven models that achieve strong performance when deployed on the physical exoskeleton. Our results reveal that no single controller is universally preferred, highlighting the necessity for personalization in gaze- driven assistive control. Our work demonstrates the utility of VR-based evaluation for accelerating the development of intuitive, safe, and personalized assistive robots.