Assessment of the occlusion technique as a means for evaluating the distraction potential of driver support systems

Driver distraction is a safety-critical issue that has been bought to greater public attention with the recent developments of more advanced driver support systems (DSS), such as navigation and collision warning. Tasks performed with such systems have the potential to distract drivers significantly from the primary task of controlling their vehicle, and this may result in an accident. Designers of DSS need to be able to assess this distraction potential in the early stages of design. In this respect, the occlusion technique and its associated measures are claimed to be reliable indicators of potential visual distraction. In particular, it has been argued that the technique provides more information concerning the likely visual demand of a system than other economical methods based on static task time, such as the 15-second rule. To investigate these assertions, a study compared results from an occlusion assessment and a road-based assessment. Sixteen experienced drivers carried out a range of tasks using two alternative user-interfaces under three conditions: statically, with full vision, statically, with restricted vision (occlusion), and whilst driving on a dual carriageway road within the UK. It was found that occlusion measures provided more information regarding the prospective visual demand of a DSS than did static task times. In particular, the resumability ratio assessed how far a task can be progressed whilst in periods with vision and without vision. It is concluded that the technique offers advantages over other methods, but requires a robust prototype for use as part of a driver-centred design process.