The Christoph Stiller Lab

Christoph Stiller’s lab conducts research on intelligent systems for automated driving, with a strong focus on data-driven perception, prediction, and decision-making in complex traffic environments. The core goal of the lab is to enable safe, efficient, and scalable autonomous vehicles that can operate reliably in real-world, highly interactive scenarios. Research topics span the full autonomy stack, including behavior modeling, multi-agent motion prediction, trajectory planning, and online high-definition (HD) map construction. The lab develops advanced machine-learning methods such as deep neural networks, transformers, generative models, inverse reinforcement learning, and optimization-based control. A key emphasis is on integrating learning-based approaches with physical constraints, navigation goals, and map priors to achieve robust and interpretable behavior. Recent work explores end-to-end and hybrid AI systems, scene-consistent prediction, controllable driving policies, and efficient simulation frameworks for large-scale traffic modeling. Applications are directly tied to real automated vehicles and have been validated in simulation benchmarks and on public roads. The research contributes to safer transportation systems, reduced traffic accidents, improved mobility efficiency, and increased public trust in autonomous driving technologies. The lab is well suited for students interested in robotics, machine learning, computer vision, or control. Ideal candidates enjoy working with real-world data, complex systems, and interdisciplinary challenges, and are motivated to combine theoretical methods with practical experimentation in autonomous driving research.