Enhancing the occlusion technique as an assessment tool for driver visual distraction

Abstract

In-vehicle information systems (IVIS) are becoming increasingly prevalent as the standard equipment of modern cars. However, there are concerns regarding the risks arising from the visual load and potential distraction caused by these systems while driving. To counteract potential compromises of driving safety the minimisation of visual demands has been acknowledged as a major goal to be achieved by safe design of the human-machine interface (HMI). In this context, the visual occlusion technique has been developed as a tool for the assessment of the in-vehicle HMI of IVIS in terms of visual demands. An unresolved issue with the standardised experimental protocol is how subjects make use of the occluded intervals and how this might affect the assessments of visual demands. We assumed that subjects would continue task performance during occluded periods leading to an underestimation of visual demands by the occlusion parameters ‘total shutter open time’ (TSOT) and the ‘occlusion index’ R. We predicted that a simple additional loading task to be performed in parallel could disrupt IVIS task performance during the occluded period leading to higher estimations of visual demands by TSOT and R. This prediction was confirmed by our data. These show significant increases in TSOT and R when the standard occlusion protocol was enhanced by a simple auditory tracking task to be performed in parallel. Furthermore, it could be demonstrated that under the condition of additional auditory tracking, TSOT and R discriminated more clearly between an ‘easy’ and a ‘difficult’ IVIS task than under the standard condition. It is concluded that amending the standard occlusion technique by imposing subjects to some cognitive load during the occlusion intervals is a viable approach to improving its sensitivity and that this aspect should be considered by future standardisation work.