Responses to and Recognition of Simultaneously-Occurring Driving Hazards using Auditory and Visual In-Vehicle Alerts

Abstract: Proper allocation of attention while driving is imperative to safety. In-vehicle alerts, meant to signal towards potential hazards, can effectively direct driver attention. However, research on in-vehicle alerts has primarily been limited to single-hazard scenarios. Scenarios in which two hazards occur simultaneously may seem unlikely – yet at least a handful of impactful real-world conditions create such a scenario. An example could include making a right turn at an intersection when there is another car turning left into the same area (hazard 1) while a pedestrian is crossing the crosswalk (hazard 2).This study aims to better understand the impact that in-vehicle alert modalities have on driver attention towards simultaneously occurring road hazards. Specifically, measuring the effect of two different in-vehicle alert modalities (auditory and visual) on driver recognition of two simultaneous hazards across two experimental drives is undertaken. Brake response time, and responses to a post-drive situation awareness question are analyzed, along with the driver’s self-reported trust in the alert system, and experience with driving and alert systems (to identify any confounding effects).Participants were randomly assigned to one of two alert conditions, auditory or visual, where they were presented with the alert modality across two different hazard scenarios. The auditory alert consisted of three short tones through the simulator speaker system, on the higher end of the range recommended by NHSTA (Jeon et al., 2022). This alert was selected to signify urgency. The visual alert consisted of an image of a red circle flashing 3 times on the simulator screen and it was confirmed in pilot testing that it did not block any of the hazards as they occurred. The semantic meanings of the two alerts were balanced and meant to alert participants to the presence of a hazard without giving them specific directions. Study findings did not support the hypothesis that auditory alerts would be more effective than visual alerts in alerting drivers to both hazards in a scenario. An analysis of variance (ANOVA) conducted on responses to the hazard observance question showed that there was not a significant effect of condition on response accuracy. An exploratory analysis of verbally indicated hazard observance by type of hazard revealed an interesting pattern in which there was one hazard that was much more likely to be recognized by participants in each scenario. This pattern was likely due to hazard salience, where the more recognized hazard may have been more salient and therefore more likely to be recognized. This is also consistent with attentional blink, in which a hazard may have been more likely to be missed while appearing alongside another, more salient, hazard. Most driving research is limited to single hazard conditions. However, recent work has demonstrated the importance of multi-hazard study (Sall & Feng, 2019; Wan & Sarter, 2022). Our study adds to this limited area and suggests that additional research should be conducted to further understand the impacts of alert modality on drivers’ hazard observance, particularly simultaneously occurring hazards.

Keywords: Human Factors in Transportation, Multiple Resource Theory, Driving Alerts and Hazards

DOI: 10.54941/ahfe1006518

Cite this paper:

​