In-Car Technology to Alert Attention-Impaired Drivers

PURPOSE: Older adults are overrepresented in fatal crashes per mile driven and attention impaired drivers show particularly elevated crash risk. We are evaluating unimodal and multi-modal alerting parameters for orienting to external targets by attention impaired drivers in different contexts. METHODS: We applied a variation of Posner‘s orienting of attention paradigm in non-demented older drivers with UFOV impairments and without impairments in a driving simulator setting. Alerting cues were presented in either a single modality or combination of modalities (visual, auditory, haptic). Several cue conditions ("valid", "invalid", "neutral", uncued) were employed, congruent with true positive alerts, false positive alerts, general alerts, and unalerted events. Following each cue, drivers discriminated the direction of a target (a Landolt square with a gap facing up or down) in the visual panorama. RESULTS: Drivers with and without UFOV impairments showed comparable response times (RTs) across the different cue types and alert modalities. Results showed that both groups benefited the most from auditory and auditory/haptic alerts, that the greatest benefit obtained when cues provided valid information, and that providing invalid spatial information is better than not cueing. CONCLUSIONS: A cognitive science approach in a driving simulator environment provides a fruitful method for examining how different alerts may influence speed of processing and attention in at risk drivers. Subsequent experiments are evaluating the effects of alerts in context, using hazard detection scenarios in the simulator setting. We are also re-evaluating the potential benefits of visual cues in augmented reality displays for older drivers in a simulator. The results should help inform the design of alerting and warning systems in real vehicles.