Urban transportation systems in China, including those in Shenzhen, currently face several prominent challenges. Traditional traffic command and control systems in megacities remain relatively inefficient and require a higher level of intelligence. Frontline traffic-control personnel are in relatively short supply, while the capacity for congestion mitigation remains limited and the resilience of urban transportation systems still needs to be strengthened. There is therefore an urgent need for transportation systems to evolve toward autonomous systems capable of intelligent decision-making, unmanned command and control, and rapid real-time response, in order to meet the future demands of smart city development.

Against this background, this research aims to develop a new intelligent and autonomous system for traffic command and risk management, with robotic agents at its core. Achieving this goal requires addressing four key technical challenges in existing systems: unreliable command and decision-making, inflexible human–robot interaction, inaccurate simulation and scenario evaluation, and inadequate supporting infrastructure.

To address these challenges, the project is organized around five major research directions, covering multimodal large-model development, human–robot interaction strategies, virtual–physical closed-loop modeling, cloud–edge–device collaborative control, and application demonstration. Ultimately, the project will establish an intelligent transportation system centered on robotic agents, supported jointly by sensing and observation capabilities, roadside control and infrastructure devices, as well as cloud-based large models and simulation-based decision-support algorithms.