I-24 MOTION Enables Generative AI for Freeway Operations
Our generative AI framework for reconstructing stop-and-go traffic waves is now published, with the dataset publicly available.
Powered by data from the TDOT I-24 MOTION testbed, our team has developed a generative AI framework that transforms conventional radar sensor data into fine-grained traffic insights - particularly for reconstructing stop-and-go traffic waves.
The peer-reviewed journal article detailing the techinical details was published in Transportation Research Part C: Emerging Technologies. The article can be found here. The dataset described in the TR-C article is now available for download!
Building on this work, we have recently received support from the NVIDIA Academic Grant Program, granting us access to 32,000 hours of A100 GPU compute for further development.