Iowa State University researchers are developing a deep learning-based system to help the Iowa Department of Transportation improve incident detection and support operator decision-making.
“There is more data than you could ever imagine coming out of this system,” said Neal Hawkins, the associate director of Iowa State’s Institute for Transportation. “We’re getting data every 20 seconds from all over the state, we’re getting high-definition camera feeds and we’re getting sensor information every minute.”
University engineers are helping the Iowa DOT by taking the data, analyzing it, making sense of it and finding ways to support improved decision-making, Hawkins said.
“The goal and outcome of TIMELI (Traffic Incident Management Enabled by Large-data Innovations) is to use emerging large-scale data analytics to reduce the number of road incidents through proactive traffic control and to minimize the impact of individual incidents that do occur through early detection, response and traffic management and control,” the Iowa State researchers wrote in a project summary.
Using TITAN X GPUs and the TensorFlow deep learning framework, the TIMELI system will be able to learn from experience and find ways to do a better job analyzing the Iowa DOT’s data streams, finding incidents and possibly even predict problems.
“Use of the system by state DOTs can reduce the duration and impacts of incidents and improve the safety of motorists, crash victims and emergency responders,” the researchers wrote.
Read more >
Related resources
- DLI course: Applications of AI for Anomaly Detection
- GTC session: How Cities and DOTs can Implement Road Safety Technology
- GTC session: Accelerate Model Training on Car Accident Data
- GTC session: Harnessing Computer Vision and Traffic Cameras for Urban Traffic Monitoring
- NGC Containers: MATLAB
- SDK: NVIDIA Fleet Command