This project develops a GPU-enabled algorithm for multi-robot path planning that combines the global evolutionary dynamic programming (EDP) with local particle swarm optimization (PSO). Notably, the algorithm runs on a cutting-edge edge computing device, the NVIDIA Jetson AGX Orin, which leverages its extensive parallel processing capabilities for faster computing. Optimal paths can be generated onboard continuously at about 0.07 seconds per path (14 Hz), demonstrating strong real-time decision-making and enabling successful obstacle avoidance for multi-robot systems.

This project tackles the challenge of delays caused by unexpected railroad crossing blockages in densely populated urban areas by integrating real-time traffic data to identify an optimal route with the shortest response time. A GPU-enabled Evolutionary Dynamic Programming (EDP) algorithm is introduced to enable rapid and reliable route planning for first responders. The algorithm's performance is validated through a series of emergency response scenarios in the City of Columbia, SC, demonstrating its ability to consistently identify optimal routes within one second.