This chapter describes the Generic Intelligent Driver Support (GIDS) project on driver-car interfaces that adapt to the workload of car drivers. 'Workload' is here defined in terms of the load on a driver's perceptual, cognitive and motor resources. An adaptive interface must ensure that these resources are not overloaded. Workload can be reduced by: (1) postponing or cancelling messages; (2) changing the modality, format and content of messages; (3) allocating tasks to a driver when workload is low, but to an automatic system when it is high. Before a driver-car interface can adapt to driver workload, a system should know when to intervene, by having an indication of that workload, and be able to estimate loads on various human processing resources. The author conducted a preliminary study, which determined the levels of visual and cognitive load as a function of the road situation, driving experience and traffic density. Twenty-four drivers drove an instrumented car along a specific route in The Netherlands, while performing one of three secondary tasks. The route had six locations where performance was analyzed. The results suggest that the visual resources of inexperienced drivers are slightly more loaded than those of experienced drivers. The author concludes that there is a need for an interface that integrates information from in-car devices, schedules them according to their priority, and adapts information presentation to current demands on the driver.
Samenvatting