Kenneth W. Gish

'Low mileage bias' and related policy implications--A cautionary note

This research examined a relationship asserted in recent literature in the field of traffic safety and injury prevention--that a significant indicator for elevated crash risk among older drivers, and potential trigger for individualized assessment at license renewal, is a low (<3000 km) annual driving distance. Sampling problems in earlier reports, in particular a reliance on self-report measures of both exposure and crash involvement, are highlighted. A pattern of misestimation for those who self-report an extremely low or extremely high number of miles driven is documented, that casts serious doubt upon the effect reported earlier. The present findings underscore the need for objective exposure measures for future analyses of this nature, and impact discussions about the feasibility of this suggested strategy to aid detection of at-risk older drivers by licensing officials.

Sensory and Cognitive Factors Affecting Automotive Head-Up Display Effectiveness

Automotive head-up displays (HUDs) present virtual images that are focused beyond the standard instrument panel distance and that appear to be superimposed on the driver’s view of the road environment. The study assessed HUD benefits and interference potential with respect to performance of in-vehicle driving tasks and response to safety-critical roadway events under various ambient light levels by older versus younger drivers. The primary baseline against which HUD performance was evaluated was a head-down display with the same message format, content, brightness, and angular size. Performance with both visual displays was compared with auditory display (AD) performance to assess channel interference. Safety-critical road events were captured on video and were presented to observers seated in a fixed-base, part-task driving simulator. Single- and dual-task trials were presented randomly within a block of trials. Measures of response accuracy and time were obtained for in-vehicle and external targets. Eye and head movements were videotaped to assess changes in scanning behavior caused by the in-vehicle displays and external targets. The analyses revealed that performances with the HUD and the head-down display were similar, whereas performance with AD was typically more accurate and faster for both in-vehicle and external responses. HUD benefits were obtained for overall response times to external targets and responses to collision avoidance warnings. Evidence for contrast interference and “cognitive capture” with HUDs and “response capture” with AD was also obtained. Implications for HUD safety and design are discussed.

Human Factors Issues Related to Use of Vision Enhancement Systems

Vision enhancement systems (VESs) are being developed to improve nighttime driving safety by providing enhanced contrast to drivers. To determine how to evaluate the potential benefits or limitations of such systems, a small-scale investigation of driver performance and behavior with a mockup VES was conducted. To maximize external validity and minimize safety risks, a field experiment was designed and conducted on a closed test track. Four younger observers (ages 26 to 36 years) and four older observers (ages 56 to 70 years) drove an instrumented vehicle and verbally reported the detection and recognition of targets while performing speed monitoring and navigation tasks. Targets near the edges of the lane of travel were detected with and without a mock-up VES and with and without headlight glare. Video recordings of the VES, road scene, driver, and data acquisition screen were obtained. Time code, distance traveled, verbal reports, and gas pedal, brake pedal, and steering wheel measurements were recorded. Results suggested that although the mockup VES provided target contrast at longer preview distances than low-beam headlights alone, the VES enhancements were not always detected by drivers because of the visual, scanning, and cognitive demands of the driving tasks. Also, older drivers were less willing to use the mock-up VES. On the basis of verbal reports, the consensus among all observers was that the VES increased curve detection distances relative to those achieved with low-beam headlights alone. The implications of the study findings for VES use, design, and evaluation are discussed.