This paper examines the difficulties associated with the application of eye movement measuring techniques to contact analogue head-up-displays (HUDs) which subtend a limited central field of view. An inherent problem lies in distinguishing between loci of attention, i.e. features in the HUD image vs. the road scene features overlaid by the HUD image, as both image and external scene coincide in space. A more basic concern arises from relating attentional allocation towards specific road scene features (inferred from direction of gaze) to sub-task performance where different sub-tasks may utilise similar sources of information. This difficulty is further exacerbated by the potential novelty of the overlaid HUD image, which in otherwise degraded viewing conditions (e.g. fog, darkness, etc.) might elicit novel strategies of visual behaviour. A task description of the functions of vision in performing driving sub-tasks is outlined, with emphasis on performance under adverse viewing conditions, and the potential impact of HUD imaging. Difficulties associated with eye movement measuring techniques can be circumvented by determining appropriate performance measures to indirectly address the visual component(s) of sub-task performance. A research programme to address key aspects highlighted in the task description is outlined. (A)
Abstract