Evaluation of Vibrotactile Warning Systems for Supporting Hazard Awareness and Safety of Distracted Pedestrians

Abstract

OCCUPATIONAL APPLICATIONS We evaluated prototype vibrotactile warning systems for supporting pedestrian awareness of approaching hazards. Mobile devices are used by people on the move for leisure and essential task activities in many work domains. While they can be a source of distraction, they also offer sensing capabilities and computing power that can detect and direct attention to approaching hazards. By presenting vibrations from devices distributed around the body, as if embedded in personal protective equipment, these systems can quickly and intuitively convey the risk of collision with a hazard and its approach direction to guide avoidance maneuvers. Results of a simulator study suggest these vibratory warnings improve response time to and avoidance of true hazards, especially when conducting distracting secondary tasks. Future developments of such systems offer promise for increasing safety for distracted pedestrians, as well as for workers in domains that impose high demand on visual and/or auditory senses. TECHNICAL ABSTRACT Background: Mobile devices can be considered essential tools for many daily activities, though there are potential safety risks associated with distraction when hazards are present (e.g., pedestrians near heavily-trafficked roads or workers in industrial settings). Yet, these devices offer onboard sensor and computing capabilities that can be leveraged to improve safety by capturing and guiding attention to approaching hazards. In domains that heavily load vision and audition, vibrotactile cues can reliably support hazard awareness and provide guidance for hazard avoidance. Purpose: We evaluated the effectiveness of prototype vibrotactile warning systems in enhancing recognition of hazards and appropriate avoidance maneuvers. Methods: Participants (n = 27) walked on a treadmill, following a virtual pedestrian path, under varied distracting task conditions. Vehicles, pedestrians, and bicyclists approached from multiple directions, representing “true” (would result in a collision if not avoided) and “false” hazards. Performance was compared with and without each of two experimental displays that presented directional vibrations via devices affixed to suspenders and to an industrial helmet. Signal detection theory and response time analyses were used to determine how well each display supported detection and avoidance of true hazards under each task condition. Results: Each vibrotactile display (suspenders- or helmet-mounted) significantly improved hazard detection in terms of hit rates and response times. Task conditions that included texting and (to a lesser extent) music negatively impacted performance, but decrements were smaller when vibrotactile displays were used. Although conditions involving the vibrotactile displays did not differ significantly in response times or signal detection theory measures, subjective ratings suggested the suspenders display was more comfortable and preferred overall. Conclusions: Vibrotactile displays improved hazard detection and avoidance; however, sensitivity in distinguishing true and false hazards could be improved. Advanced versions of these warning systems offer potential to improve hazard awareness and safety for distracted pedestrians and workers.