The AI and automation lab’s vision is to advance robotics toward robust, shared, and scalable autonomy, where robots can reliably operate alongside humans in complex, real-world environments. In self-driving laboratories today, robotics remains a critical bottleneck: manipulation is brittle, perception struggles with transparent, deformable, and fragile labware, and planning systems often fail to adapt to uncertainty, long horizons, and unexpected failures, preventing true 24/7 autonomy.

At our lab, we focus on overcoming these challenges by tightly integrating robot manipulation, computer vision, and learning-based agentic reasoning and planning systems into a unified robotic autonomy stack. We develop robots that can perceive the state of experiments, reason over tasks and constraints, and execute manipulation robustly through closed-loop control, digital twins, and data-driven learning. By combining physics-aware simulation, large-scale data collection, and adaptive planning, the lab transforms robots from scripted tools into dynamic scientific collaborators, enabling autonomous labs that run continuously, recover from errors, and dramatically accelerate discovery in chemistry, materials science, and biology.