4-7: Freeway Incident Detection and Management using Unmanned Aircraft Systems (Phase III)
|Principal Investigator||Yu Zhang, Ph.D.|
|Final Report (DOI)||Available Soon|
|Policy Brief||Available Soon|
In the last two phases of the project, the collaborative UPRM and USF research team (the research team hereafter) designed traffic data collection experiments on freeways with unmanned aerial systems (UAV with RGB and thermal cameras). The parameters of experiments included the height and speed of the drones, camera angles, congestion and non-congested traffic conditions, etc. The research team developed a learning-based object detection algorithm and evaluated the performance of the algorithm for RGB and thermal videos with different parameter settings and identified the settings with consistent high performance. In addition, the research team developed automated incident detection algorithms by identifying abnormal traffic characteristics. In Phase III of this project, the research team will focus on the validation of the algorithms developed in the previous phases and implementation matters of “Freeway Incident Detection and Management using Unmanned Aircraft Systems”. Two main tasks of the research include: (1) validate object detection algorithm and automated incident detection algorithm developed in previous phases; (2) explore the integration of UAS with traffic management center. For the first task, researchers from UPRM will focus on incident detection with CCTV video. Traffic data during the incident will be collected with drones and shared with researchers from USF. Researchers from USF will focus on applying the incident detection algorithm for the data collected during the incident. Results from both analyses will be compared and insights from the comparison will be drawn. The same procedure will be applied to analyzing traffic data from the Tampa area. For the second task, researchers from UPRM and USF will work closely with Puerto Rico DTPW and Metric Engineering, and Florida DOT District 7. By understanding the barriers and challenges of implementing emerging technologies in automatic incident detection, the research team will work on implementation recommendations.