CFLight: Enhancing Safety with Traffic Signal Control through Counterfactual Learning

2025-12-10

AI Summary

To address the prevalent efficiency-over-safety bias and limited interpretability of reinforcement learning (RL) methods in traffic signal control, this paper proposes a causal RL framework integrated with counterfactual reasoning. Methodologically, we (1) construct a counterfactual module grounded in a structural causal model (SCM) to quantify the potential impact of signal actions on crash risk; (2) design an interpretable “X-module” embedded within the policy network to explicitly encode safety constraints; and (3) introduce a near-zero-collision control paradigm. Evaluated on SUMO simulations and real-world traffic data, our approach reduces collision rates by over 35% compared to state-of-the-art RL and safe RL baselines, while simultaneously improving traffic throughput. The code and datasets are publicly released, and the framework demonstrates strong cross-scenario generalizability.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Natural Language Processing: Safety and Robustness

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Traffic accidents result in millions of injuries and fatalities globally, with a significant number occurring at intersections each year. Traffic Signal Control (TSC) is an effective strategy for enhancing safety at these urban junctures. Despite the growing popularity of Reinforcement Learning (RL) methods in optimizing TSC, these methods often prioritize driving efficiency over safety, thus failing to address the critical balance between these two aspects. Additionally, these methods usually need more interpretability. CounterFactual (CF) learning is a promising approach for various causal analysis fields. In this study, we introduce a novel framework to improve RL for safety aspects in TSC. This framework introduces a novel method based on CF learning to address the question: ``What if, when an unsafe event occurs, we backtrack to perform alternative actions, and will this unsafe event still occur in the subsequent period?'' To answer this question, we propose a new structure causal model to predict the result after executing different actions, and we propose a new CF module that integrates with additional ``X'' modules to promote safe RL practices. Our new algorithm, CFLight, which is derived from this framework, effectively tackles challenging safety events and significantly improves safety at intersections through a near-zero collision control strategy. Through extensive numerical experiments on both real-world and synthetic datasets, we demonstrate that CFLight reduces collisions and improves overall traffic performance compared to conventional RL methods and the recent safe RL model. Moreover, our method represents a generalized and safe framework for RL methods, opening possibilities for applications in other domains. The data and code are available in the github https://github.com/MJLee00/CFLight-Enhancing-Safety-with-Traffic-Signal-Control-through-Counterfactual-Learning.

Problem

• Enhances traffic signal control safety using counterfactual learning
• Addresses RL methods prioritizing efficiency over intersection safety
• Reduces collisions with near-zero accident control strategy

Innovation

• Uses counterfactual learning to enhance safety in traffic signal control
• Integrates causal models with reinforcement learning for safer intersections
• Implements near-zero collision strategy through backtracking alternative actions