Christopher Nowakowski, Dana Friedman, and Paul Green
Twenty-four drivers (12 ages 18-30, 12 ages 60-75) answered 4 implementations of hands-free cell phones while driving a simulator with 3 distinct caller ID locations: (1) a simulated cradle-mounted cell phone in the center console (a head-down implementation) with a typical auditory ring, (2) a central HUD location repeated both with and without a ring, and (3) an off-center (right) HUD location without a ring. For the center-console location, the buttons were located on the phone, but for the 2 HUD locations, the buttons were located on the steering wheel.
Drivers answered the phone very quickly (mean 2.7 seconds), often on the first ring. In general, the phone was answered more quickly for HUD implementations (2.3 seconds) than for the head-down implementation (3.8 seconds). For just over 1/3 of the drivers, suppressing the ring actually decreased (sped up) their response times. Overall, the response time increased as the curvature of the road increased (2.3, 2.5 and 3.0 seconds for 0, 3, or 9 degrees of curvature), but the effects due to momentary workload (as measured by visual demand) were less consistent.
There was little difference in driving performance between baseline driving (no incoming calls) and when the incoming calls used a HUD interface. However, for hands-free head-down interface, lane-keeping and speed-maintenance performance were degraded during incoming calls. While younger drivers showed little increase in line crossing rate (5.4 percent) using the head-down phone, the line crossing rate for older drivers was 25 percent, a value 2.5 times greater than during baseline driving. Thus, simply requiring the use of a hands-free phone (given the current implementations) will not eliminate the added crash risk while answering the phone.
|Graphical Abstract (.pdf)|||||UMTRI-2000-29 Full Report (.pdf)|
Close This Window