The Format and Presentation of Collision Warnings
Stephen Hirst Robert Graham HUSAT, Loughborough University, Leics., UK
Many road collisions are caused by the driver's failure to detect objects or other road users at an early enough stage, or by misjudgment of the obstacle's movement. Recently the development of in-vehicle collision avoidance systems (CAS) has become possible due to advances in sensor technology. The purpose of these systems is to alert the driver to potentially hazardous situations, and thereby reduce the risk of accidents caused by detection failures. However, there are significant human factors problems to be overcome before the potential of anticollision technology can be realized. Assuming that the initial uptake of CAS will be for systems exerting a minimum level of intervention, the immediate problems concern the assignment of appropriate sensory modalities for warning presentation, and the criterion to be used for warning activation.
As driving is predominantly a visuospatial task, which places its greatest load on the visual attention system, warnings presented via a visual display may themselves go undetected. Alternatively, the presence of a visual display may place demands on visual attention that compete with those required for the immediate detection of a critical incident or those needed to perform any necessary evasive action. Theoretically, these problems should be less apparent for head-up display (HUD) technology, where the image is presented in the driver's field of view such that its content can be assimilated in conjunction with foveal viewing of the driving scene ahead. The HUD image may also be presented at near infinity. Such presentation seems most applicable to high load visual driving tasks (e.g.,