C-TRAIL: A Commonsense World Framework for Trajectory Planning in Autonomous Driving

TL;DR

C-TRAIL is a framework that combines LLM-derived commonsense reasoning with a trust mechanism to improve trajectory planning safety and reliability in autonomous driving scenarios.

Contribution

It introduces a novel closed-loop cycle integrating trust-based commonsense reasoning with Monte Carlo Tree Search for autonomous driving.

Findings

C-TRAIL reduces average displacement error by 40.2%.

C-TRAIL decreases final displacement error by 51.7%.

C-TRAIL improves success rate by 16.9 percentage points.

Abstract

Trajectory planning for autonomous driving increasingly leverages large language models (LLMs) for commonsense reasoning, yet LLM outputs are inherently unreliable, posing risks in safety-critical applications. We propose C-TRAIL, a framework built on a Commonsense World that couples LLM-derived commonsense with a trust mechanism to guide trajectory planning. C-TRAIL operates through a closed-loop Recall, Plan, and Update cycle: the Recall module queries an LLM for semantic relations and quantifies their reliability via a dual-trust mechanism; the Plan module injects trust-weighted commonsense into Monte Carlo Tree Search (MCTS) through a Dirichlet trust policy; and the Update module adaptively refines trust scores and policy parameters from environmental feedback. Experiments on four simulated scenarios in Highway-env and two real-world levelXData datasets (highD, rounD) show that C-TRAIL consistently outperforms state-of-the-art baselines, reducing ADE by 40.2%, FDE by 51.7%, and improving SR by 16.9 percentage points on average. The source code is available at https://github.com/ZhihongCui/CTRAIL.