Altruistic Decision-Making for Autonomous Driving with Sparse Rewards

TL;DR

This paper introduces a new decision-making method for autonomous driving that reduces conflicts with other vehicles by incorporating altruistic behavior, verified through simulation.

Contribution

It defines a new conflict measurement (AoC) and proposes a novel altruistic decision-making approach that lowers conflict incidence compared to existing methods.

Findings

The proposed method reduces the Area of Conflict in simulations.

Altruistic decision-making improves interaction safety.

Theoretical analysis supports conflict reduction benefits.

Abstract

In order to drive effectively, a driver must be aware of how they can expect other vehicles' behaviour to be affected by their decisions, and also how they are expected to behave by other drivers. One common family of methods for addressing this problem of interaction are those based on Game Theory. Such approaches often make assumptions about leaders and followers in an interaction which can result in conflicts arising when vehicles do not agree on the hierarchy, resulting in sub-optimal behaviour. In this work we define a measurement for the incidence of conflicts, Area of Conflict (AoC), for a given interactive decision-making model. Furthermore, we propose a novel decision-making method that reduces this value compared to an existing approach for incorporating altruistic behaviour. We verify our theoretical analysis empirically using a simulated lane-change scenario.