AI-Enhanced Human-Machine Interface for Power Wheelchair Navigation: Effects of Communication Strategy and Vibrotactile Feedback Location

Mobility disabilities are among the most common disability types in the U.S., where the need to design safe powered mobility devices is increasingly important for user safety. For power wheelchair users, AI-enabled solutions have been introduced to avoid collisions. However, as control shifts from the user to automation, systems need to communicate intent clearly and maintain user trust. Auditory verbal cues can support decision-making: explanations may emphasize action-focused maneuvers or why-focused rationales. Tactile displays can also support communication through vibrations presented to the back or wrist. However, because safety-critical messages based on AI-inferred intent have not been systematically investigated and because power wheelchair users experience a different embodied interaction, there remains an opportunity to understand these factors. This pilot study explored the effects of communication strategy (action-focused, why-focused) and vibrotactile feedback location (seatback, wrist) on wheelchair performance and user perceptions. Six able-bodied participants completed wheelchair driving sessions while receiving obstacle alerts. Results indicated that seatback vibrations were associated with higher satisfaction ratings than wrist vibrations. No significant main effects were found for reaction time, decision-making accuracy, or usefulness ratings. These data can inform future studies designing intelligent wheelchair interfaces.