On-board smart vision system to support vehicle-to-infrastructure communication USDOT SBIR PhaseI&II
By: Connected Wise
During the U.S. DOT SBIR Phase-I project, Connected Wise explored an on-board AI-assisted machine vision system to support vehicle-to-infrastructure in rural areas. The system recognizes the message identifiers placed on road signs and activates the associated message while communicating with the vehicle’s on-board unit, thereby eliminating the need for wireless infrastructure in these areas. The machine vision system explored in this project integrates a deep learning framework to detect roadway entities in real-time such as traffic lights, traffic signs, vehicles, cyclists, pedestrians and crossing animals. The system also uses an affordable stereo vision system to estimate the real-time location of these roadway entities.
After completion of the Phase-I project, Connected Wise was awarded the U.S. DOT SBIR Phase-II award. The Phase-I feasibility research revealed both the advantages of the system and the challenges that need to be accounted for prior to reliable deployment. Phase-II R&D focused on improving the system’s capabilities and robustness, while investigating unique approaches that can tackle the challenges related to deployment. Connected Wise has formed strong university/industry partnerships and prepared a comprehensive R&D plan with a solid commercialization strategy for the Phase-II project. To support this collaboration, Florida High Tech Corridor matched the research funding given to the University of South Florida (USF). Connected Wise also made significant cost-sharing to show their unwavering commitment and dedication in commercializing this technology.
Phase-II R&D resulted in commercially ready technologies and products designed for different customer segments after an elaborate manufacturing process. During the investigation of commercial opportunity, many city/county traffic operation departments showed great interest in this technological solution.