Autonomous discovery of traffic laws with AI traffic scientists

2026-07-01

AI Summary

Traditional approaches to uncovering traffic regularities rely heavily on expert intuition and struggle to automatically extract generalizable patterns from complex urban data. This work proposes TrafficSci, an AI system grounded in a multi-agent architecture that formalizes scientific discovery as an auditable, iterative process integrating evidence scoping, hypothesis generation and critique, hybrid observational–interventional validation, and cross-scale analysis. TrafficSci represents the first demonstration of autonomous, generalizable law discovery by AI in complex urban traffic systems. In four case studies, it successfully reproduces three established traffic laws and consistently identifies—across trajectory datasets from eight cities—a previously unreported intrinsic temporal memory scale underlying driving behavior.

Abstract

Universal traffic laws describe recurrent patterns in congestion, mobility and driving behavior across cities, providing a scientific basis for transportation planning, management and control. Their discovery, however, remains expert-driven, requiring candidate regularities to be identified from heterogeneous observational evidence or validated through intervention experiments. Although autonomous artificial intelligence (AI) systems have advanced scientific discovery in controlled laboratory settings, extending them to complex transportation domains remains a challenge. Here we present TrafficSci, an agentic AI system that formulates traffic-law discovery as an iterative, auditable workflow integrating evidence scoping, critic-judge hypothesis induction, and observational-interventional validation. Across four case studies spanning population, network, control and trajectory scales, TrafficSci autonomously rediscovers three established traffic laws and identifies an unreported intrinsic temporal memory scale in urban driving behavior, statistically consistent across eight cities and two trajectory datasets. TrafficSci provides a route for extending AI-driven scientific discovery from controlled domains to complex urban systems.

Problem

• autonomous discovery
• traffic laws
• AI-driven scientific discovery
• urban systems
• complex transportation domains

Innovation

• autonomous scientific discovery
• AI agent
• traffic laws
• observational-interventional validation
• urban mobility