Inattentional blindness to unexpected hazard in augmented reality head-up display assisted driving: The impact of the relative position between stimulus and augmented graph.

Objective: An augmented reality head-up display (AR-HUD) is a promising technology in assisted driving. It provides additional information in the driving environment. However, considering the registration problem related to the limitations of interactive technology, we suspect that an AR-HUD may not be able to recognize unpredictable stimuli in a timely manner, inducing inattentional blindness to these non-augmented stimuli. Actually, non-augmented stimuli may accidentally have a brief superimposition to AR graphics. This condition may also influence the rate of inattentional blindness accordingly. Thus, this study examined the problem of inattentional blindness in AR-HUD systems that may result from the immaturity of AR technology.

Method: We investigated the impact of AR graphic position (peripheral AOI v.s. central AOI) and the relative position of the AR graphic on unpredictable stimuli (on-HUD hazard v.s. off-HUD hazard) on the occurrence of inattentional blindness. Thirty Participants watched an AR-augmented driving video that included four augmented conditions. Participants were instructed to respond to four critical events (speeding, running of red lights, unexpected pedestrians or motorcycles). The rate of inattentional blindness and response time were recorded. We only analyzed data on unexpected pedestrian and motorcycle incidents.

Results: The relative position of the AR graphic on unpredictable stimuli and AR graphic positions significantly affected the rate of inattentional blindness and response time. Drivers had a higher rate of inattentional blindness to the unpredictable stimulus briefly superimposed on the AR graphic (i.e., on-HUD hazard) in the peripheral visual field (i.e., peripheral AOI). Also, drivers exhibited a higher rate of inattentional blindness to the unpredictable stimuli outside the AR graphic (i.e., off-HUD hazard) in the central visual field (i.e., central AOI).

Conclusion: The study is expected to be beneficial for furthering the design of an AR-HUD-assisted system to reduce inattentional blindness in driving. Our results found that in the peripheral visual field, unpredictable stimuli accidentally superimposed on the AR graphic (i.e., on-HUD hazard) lead to a higher probability of ignoring the accidental events and seemed to require a longer response time for drivers. This study illustrated that inattentional blindness to non-augmented stimuli is also influenced by the AR graphic position when AR technology fails to augment them in a timely manner. An important recommendation emerging from this work is to consider the design of AR graphics according to the AR graphic positions and stimulus types to reduce the occurrence of inattentional blindness.