This project utilizes Event Streaming Platform (ESP), a central environment for ingesting and outputting any transportation-related data in near real-time, to deploy these proofs-of-concept.  Etch, LTD put together a team of professionals with Trihydro and moove.ai, to successfully complete this project.

Incident Detection and Response Initiation

Etch was able to create a simulated environment to receive air bag deployment alerts directly from vehicle manufacturers and automatically provide crash location and severity information to emergency responders. This offers ODOT or emergency personnel an enhanced capability to detect, react, and respond in areas where traffic cameras and roadside units are not available.

Crash Prediction and Proactive Mitigation

This included real-time and static roadway data from disparate systems such as: speed, traffic volumes, weather, pavement conditions, work zones, average daily traffic volumes, transportation systems management, and operations studies. These are compared to historical data to automatically infer scenarios that pose a higher probability of crashes occurring. By ingesting this data and adding a trained machine learning model in the ESP, ODOT can systematically provide mitigation recommendations to the Traffic Management Center operators to reduce the risk of crashes.