An Energy-efficient and Self-diagnostic Portable Edge-Computing Platform for Traffic Monitoring and Safety

This project aims to develop an energy-efficient and self-diagnostic portable traffic monitoring system utilizing AI-based edge computing technologies. The system, named ThermoVision Edge, addresses specific technical challenges encountered in Intelligent Transportation System (ITS) applications, including comprehensive sensor integration, efficient data processing and information extraction, sustainability in remote areas, customizability, retrofitting existing infrastructure, and accommodating varying environmental and weather conditions. Designed for low power consumption, it can be supplemented with solar energy and features a self-diagnostic function to ensure reliability. This system is engineered to provide in-depth traffic insights with minimal human oversight, aiming to enhance transportation infrastructure management. In Phase I, the evaluation of prototype edge-computing devices and the creation of an energy-efficient hybrid camera prototype for vehicle and micromobility classification using both visible and thermal video data, showcased promising results. Phase II will focus on fully developing, deploying, and commercializing the ThermoVision Edge system, which utilizes deep-learning models to classify vehicles, detect Vulnerable Road Users (VRU), identify stopped vehicles, recognize conflict events, and identify pavement and environmental conditions. This will be achieved by successfully demonstrating its operation and evaluating its performance in three states.