Modeling Driver Behavior at the On-set of a Yellow Phase Transition

This report presents five research efforts that are being published in various journals. The first research effort conducts two experiments to provide partial validation for the concept of a system to warn the potential victims of red-light violators. As conceived, the warning system would entail sensors to detect approaching vehicles that are unlikely or incapable of stopping at red traffic signals. The system would then use flashing signs and lights to warn drivers on the intersection approaches that are in the green phase. Several human factors issues are identified that would need to be addressed before such a system could be deployed. The present experiments addressed one of these issues—would a sufficient number of drivers respond to such a warning in a way that would avoid collision with a red-light violator? The results suggest that the majority of drivers will brake sharply in response to a conspicuous warning, at least for the case where no other vehicles are preceding or following the potential victim. Comparisons between field test and driving simulator response measures are discussed, as are the implications for further research. The second effort characterizes driver behavior at the onset of a yellow-phase transition on highspeed signalized intersection approaches using field data gathered from 60 test subjects (approximately balanced in gender and age). The driver stopping/running decisions are analyzed for five trigger distances, measured from the vehicle position to the stop bar when the yellow phase starts, as drivers approach the intersection at a speed of 72 km/h (45 mi/h). The study demonstrates that the probability of stopping varies from 9 percent at a time to intersection of 1.6 s to approximately 100 percent at a time to intersection of 5.5 s. The study also demonstrates an increase in the probability of running for male drivers when compared to female drivers. This difference increases as the trigger distance decreases. Furthermore, the study demonstrates that drivers 65 years of age and older are significantly less likely to clear the intersection at short yellow-phase trigger distances when compared to other age groups. The dilemma zone for the less than 40 year age group is found to range from 1.85 to 3.9 s while the dilemma zone for the greater than 70 year age group is found to range from 1.5 to 3.2 s. Dilemma zone boundaries (distances where 10 to 90 percent of the vehicles stop) are derived and uncertainty zones for different age groups are also developed. The third effort used data from a field test on 60 test participants to characterize driver brake perception-reaction times at the onset of an amber phase actuation at a high-speed signalized intersection approach. The study demonstrates that the 1.0 s, 85 percentile perception-reaction time that is recommended in traffic signal design procedures is valid and consistent with field observations. Furthermore, the study demonstrates that brake perception-reaction times are impacted by the vehicle’s time-to-intersection at the onset of a phase transition. Consequently, the study recommends that time-to-intersection in the range of 2.2 to 4.4 s be considered in the characterization of brake perception-reaction time behavior. Finally, the study demonstrates that either a lognormal or beta distribution is sufficient for modeling brake perception-reaction time. The fourth effort characterizes driver stopping behavior at the onset of a yellow-phase on highspeed signalized intersection approaches using controlled field data gathered from 60 test subjects using an in-vehicle Differential Global Positioning System (DGPS). A total of 745 data records were available for analysis for all drivers who stopped at the onset of the yellow-phase, ranging from a minimum time to the stop bar (TTS) of 1.34 s to a maximum of 6.19 s. Statistical