SCC-Planning: Pedestrian Safe and Secure Communities with Ambient Machine Vision

点击翻译按钮获取中文摘要

This project with the University of North Carolina at Charlotte in cooperation with the Charlotte-Mecklenburg counties addresses community challenges of pedestrian safety and community policing, building on advances in cyber-physical systems (CPS). As communities adopt technologies such as vision-based traffic cameras and smart traffic signs at intersections, the data from these technologies possess traces of the activity within a community of which a few might need a response because of risk to individual and public safety or suggest a local police response. Such technologies may provide a more accurate community-wide operational picture. With this data communities can have a better understanding of itself and within established law and custom will better serve and protect individuals and the public at large. This planning grant will enable community planners, local government, and businesses along with technologists, urban planners and traffic engineers to explore the potential of these emerging technologies for improving the quality of life of a community.This planning grant will leverage research in CPS, big data, and urban transportation planning to provide new capabilities for community engagement. It will draw upon technologies from computer vision, machine learning, edge computing, and generally CPS and the Internet of Things. This will set the stage for designing edge computing systems for ambient vision processing at city street intersections with cooperative processing over the entire edge network in a city. The project will advance knowledge of pedestrian and driver behaviors and models, specifically in urban transportation settings. It will enable the study and characterization of driver behaviors for driver-in-the-loop traffic control system. The planned extensive community engagement will facilitate ascertaining community goals and concerns, especially regarding privacy and transportation mobility planning in future community deployment of these proposed technologies.