Shared-Control HMI for Tactile-First Traversal Offline Counterfactual Evaluation with Haptic Safety Projection

Area of Science:

• Robotics
• Human-Computer Interaction
• Control Systems

Background:

• Supervising robots in confined, uncertain environments requires operator intervention without micromanagement.
• Tactile-first robotic traversal presents unique challenges for human oversight.

Purpose of the Study:

• To present and evaluate a human-machine interface (HMI) that blends operator commands with safety-constrained autonomy.
• To surface risk through synthesized predictive haptic alerts for improved operator awareness.

Main Methods:

• Offline, log-driven replay of 660 trials for counterfactual evaluation of the HMI.
• Development of a shared-control system integrating operator input with safety-aware autonomy.
• Implementation of synthesized haptic alerts to predict safety-critical events.

Main Results:

• Consistent improvements observed: decreased predicted collisions, increased minimum clearance, and enhanced traversal efficiency.
• Reduced operator-autonomy disagreement, leading to smoother control and fewer heading reversals.
• Haptic alerts accurately anticipated safety-critical events with positive lead time, demonstrating high informativeness.

Conclusions:

• Shared-control blending with tactile-first autonomy enhances safety, efficiency, and assurance in robot operation.
• The developed HMI provides a practical pathway for safe, low-overhead operator supervision in vision-denied environments.
• The study contributes a novel method, metrics, empirical results, and artifacts for tactile-first HMI research.