Merging into strings of cooperative-adaptive cruise-control vehicles

Abstract

Cooperative-adaptive cruise-control (CACC) offers the potential to increase roadway capacity by decreasing the size of gaps between vehicles. However, smaller gaps may reduce drivers’ ability to safely enter CACC strings. Two driving simulator experiments explored merging in CACC strings (also known as CACC platoons). In Experiment 1, drivers attempted to merge into a continuous stream of simulated CACC vehicles using either a CACC vehicle with merge assist, a CACC vehicle without merge assist, or a manually driven vehicle. All merges made using merge assist were successful. In contrast, one-third of CACC drivers without merge assist and half of manual drivers were involved in a collision during their first merge attempt. The crash rate appeared to be largely due to participants’ expectation that CACC vehicles would create a larger gap to accommodate their merging vehicle. Experiment 2 explored the validity of this expectation. Participants encountered two merge events while traveling in a CACC string: a successful merge that occurred immediately in front of their vehicle and an unsuccessful merge attempt that resulted in a crash in their travel lane. Responses to merging vehicles were influenced by the gap size their vehicle maintained, but not by their preferred following distance. The vast majority of participants traveling in a CACC string did not create larger gaps during the typical merge, and instead only took over longitudinal control when in immediate danger of a collision. The finding suggests that features, such as merge assist, could help ensure the safety of vehicles entering CACC strings.