Cooperative Adaptive Cruise Control Human Factors Study: Experiment 4—Preferred Following Distance and Performance in An Emergency Event

CHAPTER 4. DISCUSSION

There were many goals of this study. The first part of this study sought to determine the typical range of the shortest time gap that drivers feel comfortable following. Participants drove in two scenarios where they were asked to follow a lead vehicle at a comfortable distance and at a minimally safe distance. The median gap from the minimum safe drive was used to classify participants in the second study as near or far preference followers.

In the second part, participants first drove in the two scenarios used in part one. The mean following distance from the minimum safe drive was used to determine participants’ preferred following gap. Those who drove with a shorter gap than the median value from the first study were determined to be close (or near) followers, and those with longer time gaps were determined to be far followers.

Next, participants drove in a scenario with CACC engaged with either a near (0.6 s) or far (1.1 s) time gap. During that drive, a vehicle merged directly in front of the participant. Later, a vehicle out of direct line of sight crashed, which required the participant to take action to avoid collision. Driver performance during these two events and perceived workload were assessed.

As one might expect, drivers’ perceived workload (as assessed by the NASA-TLX) varied by location at which it was administered. Not surprisingly, workload was the greatest after the final crash event. However, no significant variation between the workload after the initial merge event and a relaxed cruise time period was found. This provides evidence that participants felt that the CACC system was adequately able to allow space for a merging vehicle.

At the first vehicle merge event, only three people depressed the brake. While this highlights an overall trust in the system, foot hovering behavior highlighted that those people following at the near CACC setting were prepared to override the system if necessary as the other vehicle merged. This was independent of preferred following distance. That is, those participants who preferred to follow at closer time gaps were as likely to hover their foot over the brake pedal in anticipation of potentially overriding the CACC system as those who preferred to follow at longer time gaps.

During the emergency event, those participants assigned to drive at the closer following distance both reacted faster and had more crashes. Given that participants following closer had physically less distance to react to avoid a collision, it is not surprising that more of these participants experienced a collision.

Throughout the study, participants’ following distance preference did not affect performance. In other words, participants’ abilities may not necessarily reflect their following preferences. This is a promising finding for widespread implementation of CACC. While overall comfort level may vary across drivers, these findings support the idea that performance will depend more on overall CACC following distance settings than with drivers’ personal preferences. This will allow CACC systems to implement a single following distance gap (or set of gaps based on vehicle physics). These results also highlight the need to implement well-designed human factors-based systems that clearly indicate to drivers when it is necessary to take over control of the vehicle. Without such measures, it is possible that CACC implementation may not result in roadway safety improvements.