Assessing the Effectiveness of Connected Vehicle Technologies based on Driving Simulator Experiments
Connected vehicle (CV) technology is expected to reduce crashes and improve roadway safety, though the specific effects will depend on crash scenarios. Differing crash scenarios may be associated with variations in a driver’s perception and interpretation, and lead to different driver behaviors and evasive strategies in response to hazards. Researchers at the University of Central Florida aimed to assess this variability in the performance of CV technology between different crash scenarios using a driving simulator experiment with two types of CV applications:
- Forward collision warning (FCW) which used the warning message “Keep Distance!”
- Pedestrian-to-vehicle (P2V) which provided a “Slow Down! Pedestrian Crossing!” message to the driver
A total of 49 licensed drivers completed the experiment, which consisted of driving simulated scenarios both with and without the applications providing safety warnings. FCW was tested in four rear-end crash scenarios, and P2V was tested in three pedestrian crash scenarios. The four rear-end crash scenarios provided the driver with different driving hints regarding the leading vehicle’s behavior and included a distraction task. The three pedestrian crash scenarios involved unobservable pedestrians due to obstruction or darkness, and unpredictable pedestrians that suddenly changed their posture and turned into the road.
Two variables were used to quantify the safety benefit of FCW, namely the collision rate, and the minimum modified time to collision (MMTTC) based on the vehicles continuing to travel at current speeds and acceleration levels. To quantify the safety benefit of the P2V warning, the collision rate, impact velocity, and post-encroachment time (PET) were used. The PET represents the time difference between when a conflicting road user leaves a location and the arrival of the other conflicting road user, with a higher value indicating a larger margin of safety.
The simulation results showed that FCW technology could reduce the probability of a crash, increased the MMTTC, and reduced the driver brake reaction time.
- The reduction in rear-end crashes due to FCW varied from 56.6 to 69.8 percent in four tested scenarios.
- FCW technology increased MMTTC by up to 53 percent (from 0.89 s to 1.36 s), depending on the scenario.
- FCW technology reduced the brake reaction time by up to 23 percent (from 1.60 s to 1.23 s).
Similarly, the simulation results showed the effectiveness of P2V technology in reducing the probability of a crash, and increasing the PET.
- P2V warning technology reduced pedestrian crashes by 89.2 to 97.2 percent in the three tested scenarios.
- P2V warning messages resulted in lower driver brake reaction times, by up to 97 percent (from 1.07 s to 0.03 s), depending on the scenario.
- The warning was less effective in the scenario when the vehicle was approaching a pedestrian hidden behind a parked car. The PET increased by 2.73 s compared with increases of 3.94 s and 4.34 s in the other scenarios.