CuraLight: Debate-Guided Data Curation for LLM-Centered Traffic Signal Control

TL;DR

CuraLight introduces a novel framework combining reinforcement learning and large language models with debate-based data curation to improve traffic signal control, achieving better efficiency and interpretability in diverse real-world scenarios.

Contribution

The paper presents a new RL-assisted, debate-guided data curation approach for fine-tuning LLMs in traffic signal control, enhancing scalability and interpretability.

Findings

CuraLight reduces average travel time by 5.34% in real-world networks.

It decreases average queue length by 5.14%.

It lowers average waiting time by 7.02%.

Abstract

Traffic signal control (TSC) is a core component of intelligent transportation systems (ITS), aiming to reduce congestion, emissions, and travel time. Recent approaches based on reinforcement learning (RL) and large language models (LLMs) have improved adaptivity, but still suffer from limited interpretability, insufficient interaction data, and weak generalization to heterogeneous intersections. This paper proposes CuraLight, an LLM-centered framework where an RL agent assists the fine-tuning of an LLM-based traffic signal controller. The RL agent explores traffic environments and generates high-quality interaction trajectories, which are converted into prompt-response pairs for imitation fine-tuning. A multi-LLM ensemble deliberation system further evaluates candidate signal timing actions through structured debate, providing preference-aware supervision signals for training. Experiments conducted in SUMO across heterogeneous real-world networks from Jinan, Hangzhou, and Yizhuang demonstrate that CuraLight consistently outperforms state-of-the-art baselines, reducing average travel time by 5.34 percent, average queue length by 5.14 percent, and average waiting time by 7.02 percent. The results highlight the effectiveness of combining RL-assisted exploration with deliberation-based data curation for scalable and interpretable traffic signal control.