Welcome to the Safe and Intelligent Autonomy (SIA) Lab! SIA Lab is part of the Department of Aeronautics and Astronautics at Stanford. The central question our lab aims to answer is how can we enable robots to leverage the capabilities offered by modern ML and AI methods while maintaining safety guarantees?
Towards this goal, we develop theory and computational tools for robot safety verification (e.g., neural reachability methods for high-dimensional and black-box systems) and integrate them into different stages of the learning lifecycle. This includes methods to provide safety assurances during the design and training phase (e.g., via safety-guided data collection and learning correct-by-construction robot policies); continuous safety monitoring during test and deployment (e.g., through uncertainty quantification, anomaly detection, and safety shielding); and dynamically adapting to new safety risks during the robot’s lifecycle (e.g., via developing new stress-testing methods that expose hidden AI failure modes, trace their root causes, and enable targeted fixes). Our work is applied in a variety of domains, such as flying vehicles (quadrotors, eVTOLs), quadrupeds, autonomous cars, space exploration, and manipulation.
Please see this talk for a nice summary of the latest research directions in our group.
We are looking for undergraduate and graduate students to join our research group. If you are interested in robotics, control, and/or machine learning, consider joining us!
News
[August 2025] New paper on Safety-Aware Imitation Learning via MPC-Guided Disturbance Injection is out on arXiv.
[July 2025] New paper on Safety Evaluation of Motion Plans Using Trajectory Predictors as Forward Reachable Set Estimators is out on arXiv.
[June 2025] New paper on Safe and Performant Deployment of Autonomous Systems via Model Predictive Control and Hamilton-Jacobi Reachability Analysis is out on arXiv.
[June 2025] New paper From Perception Logs to Failure Modes: Language-Driven Semantic Clustering of Failures for Robot Safety is out on arXiv.
[June 2025] Congratulations Javier, Luke, and Umut, for your paper acceptance to IEEE Robotics and Automation Letters.
[June 2025] Congratulations Manan and Aditya, for your paper acceptance to International Conference on Machine Learning (ICML), 2025.
[May 2025] New paper on Reachability Barrier Networks: Learning Hamilton-Jacobi Solutions for Smooth and Flexible Control Barrier Functions is out on arXiv.
[April 2025] Congratulations Zeyuan and Le, for your paper acceptance to Robotics Sciences and Systems (RSS), 2025.
[January 2025] Five papers from SIA Lab have been accepted to ICRA 2025. Congratulations, Umut, Kaustav, Vamsi, Aditya, and Zeyuan!