Cooperative Autonomous Vehicles: An Investigation of the Drop-Off Problem

Abstract

Autonomous vehicles (AVs) will revolutionise city design, and radically affect how people move. The scenario of AVs taking users to their destinations and then conveniently dropping them off is an obvious advantage, reducing the need for drivers to waste time car parking. However, it is not so clear how AVs need to behave when dropping off passengers at busy drop-off areas, since there are a number of possible options, from simply waiting for a spot at the drop off area, circulating till a slot is available at a drop-off area to actually parking further, depending on the situation at the drop-off area and passenger preferences. In this paper, we propose and investigate a cooperative drop-off (CDO) approach. CDO can handle the unpredictable circumstances of a drop-off situation using vehicle cooperation and makes rational collective decisions with the goal of minimizing the arrival time of passengers. Comprehensive simulation of a range of aspects is used to assess and examine different factors affecting the CDO approach, and it is shown to provide reasonable performance under different conditions.