This project leverages connected intelligent transportation technologies to enhance the safety of vulnerable road users (VRUs) within evolving urban road systems. This project targets three fundamental barriers: the lack of high-fidelity experimental platforms, localization failures in GNSS-denied areas, and insufficient trajectory prediction for heterogeneous agents. Our goal is to develop an integrated cooperative system that combines virtual reality, wireless communication, and physics-informed learning. To achieve this, we develop Sky-Drive, a distributed multi -agent simulation platform that enables human -in-the-loop interaction testing. Furthermore, the project implements a cellular Vehicle-to-Everything (C-V2X)-based cooperative localization framework to ensure lane-level accuracy and a kinematics-aware multigraph attention network for precise motion forecasting. The output of this project provides a comprehensive technological foundation for enabling the safe and equitable coexistence of autonomous vehicles (AVs) and VRUs in mixed traffic environments.

Infrastructure Assets: Highway Assets, ITS, Other Highway Assets, Pavement
Resource Types: Research Report
Capabilities: Data & Information Systems, Tools & Technology
Management Processes: Performance Based Planning & Programming, Performance Reporting & Communication, Risk Management, Strategic Direction, Target Setting