A study of collaborative influence mechanisms for highway convoy driving

Convoy driving on highways is a desirable behavior which reduces the risk of highway accidents and makes traffic faster and more fluent. Recent technologies, such as intelligent cruise control devices explicitly facilitate convoy driving by providing a fully automated means for following the previous vehicle. Participating in a convoy, however, requires compromises from the vehicles, such as slowing down to the speed of the lead vehicle; thus many drivers choose not to join any convoy. Collaborative convoy driving systems, based on vehicle-to-vehicle communication, promise to deliver means for the vehicles to influence the speed of the convoy, thus improving its utility. We discuss the mechanisms of convoy participation, including the decision to join and leave the convoy, and the mechanisms through which the vehicles can influence the convoy speed. In an experimental study, we compare three influence mechanisms: the “adapt speed to the leader” mechanism used by human drivers and intelligent cruise control systems and two collaborative influence mechanisms which require vehicle to vehicle communication. We show that the collaborative cruise control methods deliver better macroscopic performance measures: more vehicles participating in convoys, higher average speed and lower number of overtakings.