Calibrating Urban Traffic Simulation from Sparse Road Observations via Genetic Optimization
2026-06-02 • Artificial Intelligence
Artificial IntelligenceComputers and SocietyNeural and Evolutionary Computing
AI summaryⓘ
The authors created a method to improve city traffic simulations even when detailed traffic and job location data are limited. They use a genetic algorithm to adjust traffic models based on a small set of known traffic measurements without needing exact employment data. Testing their method in Greensboro, North Carolina, showed that simulated traffic matched real traffic well and predicted traffic on roads they didn't use to train the model. Their approach helps make traffic simulations more practical and usable in many cities with minimal data.
urban traffic simulationgenetic algorithmtraffic calibrationemployment distributionSUMO simulationtraffic flowcommuter trafficinfrastructure planningdata sparsity
Authors
Hunter Sawyer, Jesse Roberts, Simon Matei
Abstract
Urban traffic simulation is a critical tool for infrastructure planning, including the placement of electric vehicle charging stations. However, realistic traffic simulation across many cities is hindered by two fundamental data limitations: detailed real-world traffic measurements are available for only a small fraction of road segments in most cities, and employment distribution data critical for modeling commuter traffic is rarely available at the resolution needed for simulation. This paper presents a genetic algorithm-based framework that directly addresses both limitations, calibrating urban traffic simulations from sparse road observations without requiring detailed job location data. Using the SUMO traffic simulation platform for Greensboro, North Carolina, our approach optimizes job distributions and gate-traffic parameters to align simulated traffic with a small sample of roads with known traffic-flow rates. We demonstrate that this approach produces simulated traffic that correlates well with real-world measurements, generalizes to road segments withheld from training, and produces job distributions that show promising qualitative agreement with census employment data despite never directly training on that employment data. This work demonstrates that realistic urban traffic simulation can be achieved from minimal real-world observations, offering a scalable and data-light approach to simulation calibration that reduces the barrier to deploying traffic models across diverse cities.