This report (1) identifies measures of the safety and ease of use of driver information systems, (2) describes test protocols for assessing safety and ease of use, and (3) identifies levels of acceptance. Only the driver interface is considered, not system safety considerations. Two protocols are described: an initial on-road test to assess the basic interface, and follow-on surveys at driver licensing offices after only small changes are made to the interface. The on-road test involves use of an instrumented car. From the data collected, measures of the standard deviation of lane position, mean speed, speed variance, the number and duration of eye fixations, and interface-specific performance measures (e.g., the number of turn error) can be obtained. For each measure, three levels of acceptance are specified: best expected, desired/planned, and worst case. The measures listed above should be viewed as suggestions only. Normative data on driver performance are lacking, and the validity of the test protocols has yet to be established. There are also concerns about these procedures not being cost effective.
Abstract