Signal Optimal Scheduling of Road Maintenance Jobs Considering Impact on Traffic Flows

Summary

A study addresses network-level road maintenance planning, a problem that requires repeatedly computing equilibrium traffic flows whenever road capacity is reduced. Equilibrium traffic assignment models are already well established, but embedding them directly and repeatedly within a maintenance scheduling problem quickly becomes computationally prohibitive. The authors investigate data-driven surrogate models that approximate equilibrium link flows straight from origin-destination demand, using optimization-based equilibrium solutions as the ground truth for training. They validate the approach with a real-world case study built on traffic data from the Newark, New Jersey area. The authors conclude that the surrogate model can serve as a scalable building block within future maintenance scheduling frameworks.