The goal of my research program is to develop automated decision-making methods for cyber-physical systems to achieve optimal and robust performances. Such systems include multiple, heterogeneous entities (humans, robots, smart devices, passive components, etc.) and occur at widely varying spatial and temporal scales from controlled micro-bio environments to large manufacturing enterprises. The methods have to be novel to address unsolved challenges, theoretically sound to provide performance guarantees, and both computationally efficient and easily realizable to be practically useful. These requirements lead to a cyclic approach involving the tight coupling of mathematical and physical modeling of system dynamics, model-based stochastic planning and control, and system behavior data-driven knowledge discovery leading to continuous model refinement. Please visit the Projects and Publications pages for details on the specific methods and the application domains.