IoT Thrust Seminar | Data-driven Human Cyber-Physical Systems for Smart Cities

Today, with billions of devices online, the generated large-scale multi-modal data presents a unique opportunity to identify and address real-world challenges (e.g., efficiency and sustainability) in smart cities. My research focuses on creating a positive impact for urban residents through data-driven models and a Cyber-Physical System (CPS) approach, which integrates sensing, computation & prediction, decision-making, and networking into physical infrastructure. In this talk, I will use urban logistics as a concrete scenario to show how we address real-world challenges by bridging Human, AI, and CPS. First, I will show how we can perceive continuous human behaviors at a large scale by designing low-cost mobile sensing systems and generalizable AI models. Then, I will show how to model fine-grained behaviors to improve logistics systems, e.g., reducing energy consumption and enhancing operational efficiency. Finally, I will conclude this talk with key insights and discuss future directions.