This fact sheet highlights the research's team holistic framework to address the challenges in large-scale predictive traffic incident management (TIM). The research can be summarized as interconnecting subtask models that accomplish the following:
 

•     Predict traffic speed.
•     Detect traffic anomalies.
•     Approximate traffic flow physics.
•     Control traffic.
•     Estimate network benefits (e.g., mobility, safety, and energy use).
   

The researchers want to predict nonrecurrent traffic conditions in large-scale networks up to 30 min ahead of the earliest time an incident is reported and proactively recommend real-time operational management strategies.

Recommended citation: Federal Highway Administration, Predictive Real-Time Traffic Management in Large-Scale Networks Using Model-Based Artificial Intelligence (Washington, DC: 2023) https://doi.org/10.21949/1521439