Emerging cooperation on the road by myopic local interactions

TL;DR

This paper presents a distributed algorithm enabling autonomous vehicles with limited memory and sensing to cooperatively reach their destinations on multi-lane roads without collisions, despite minimal communication.

Contribution

It introduces a novel distributed approach for autonomous vehicles with limited capabilities to coordinate effectively on multi-lane roads.

Findings

Vehicles reach their destinations without collisions

The algorithm operates efficiently with limited memory and sensing

Coordination is achieved without inter-vehicle communication

Abstract

We study a combinatorial problem inspired by the following scenario: fully autonomous vehicles drive on a multi-lane ($m \geq 2$) road. Each vehicle heads to its own destination and is allowed to exit the road only through a single designated off-ramp lane. However, an individual vehicle has a severely limited memory and sensing capabilities, and, moreover, does not communicate with its peers. In this work we present a distributed algorithm that, nonetheless, allows vehicles to get to the desired lane without collisions and in timely manner.