My PhD dissertation, completed in the department of Cognitive Robotics (CoR) at Delft University of Technology under the supervision of Prof. Javier Alonso-Mora and Prof. Laura Ferranti.
Robots that operate among humans and other robots must reason about how their own actions influence the decisions of those around them. This dissertation develops game-theoretic methods for motion planning in such multi-agent settings, where the objectives of interacting agents are coupled and, in many cases, only partially known. The presented contributions span efficient solvers for dynamic games, methods for inferring other agents’ objectives from observed behavior, and game-theoretic planning techniques that remain robust under uncertainty. Throughout, the developed methods are grounded in real-world applications such as autonomous driving and mobile robot navigation.