Evaluation of Large Language Models for Decision Making in Autonomous Driving

TL;DR

This paper evaluates the spatial recognition and decision-making capabilities of large language models in autonomous driving and demonstrates a vehicle control system powered by LLMs.

Contribution

It provides a quantitative assessment of LLMs' abilities in autonomous driving tasks and presents a prototype vehicle system utilizing LLMs for decision making.

Findings

LLMs can recognize spatial information relevant to driving.

LLMs can make basic driving decisions based on input prompts.

Prototype system successfully navigates in controlled scenarios.

Abstract

Various methods have been proposed for utilizing Large Language Models (LLMs) in autonomous driving. One strategy of using LLMs for autonomous driving involves inputting surrounding objects as text prompts to the LLMs, along with their coordinate and velocity information, and then outputting the subsequent movements of the vehicle. When using LLMs for such purposes, capabilities such as spatial recognition and planning are essential. In particular, two foundational capabilities are required: (1) spatial-aware decision making, which is the ability to recognize space from coordinate information and make decisions to avoid collisions, and (2) the ability to adhere to traffic rules. However, quantitative research has not been conducted on how accurately different types of LLMs can handle these problems. In this study, we quantitatively evaluated these two abilities of LLMs in the context of autonomous driving. Furthermore, to conduct a Proof of Concept (POC) for the feasibility of implementing these abilities in actual vehicles, we developed a system that uses LLMs to drive a vehicle.