Scalable and Sustainable AI Solutions for Complete Streets Data Creation and Maintenance with Off-the-shelf Dash Cams

The creation of a complete street data set is the first step in enabling new services for pedestrians, bicyclists, scooter users, EV drivers, delivery/ride share drivers to reduce traffic congestion and illegal parking, save drivers' travel time and gas, improve travel safety and comfort, optimize planning for extending connectivity, equity, and reachability. This project will study AI techniques for efficiently creating and maintaining a complete street data set via analyzing the street videos collected by off-the-shelf dash cams. The objectives of this research are to (1) study AI technologies for creating the complete street data from street videos, (2) develop routing algorithms and mobile app for data maintenance, (3) design secure and privacy protection data storage schemes, and (4) develop a cross validation system to demonstrate the collected data's accuracy and design APIs for sharing the data in standard formats.