About Me
I am a Robotics Researcher working at the intersection of robotics, machine learning, and computer vision to advance robot manipulation capabilities. My research focuses on reducing the data requirements for robot learning, enabling robots to acquire new skills from minimal demonstration data while performing reliably in real-world environments.
My journey in robot learning began with my PhD at Imperial College London's Robot Learning Lab under Dr. Edward Johns, where I focused on leveraging inductive biases to enable robots to learn new manipulation tasks from less than a minute of human demonstration time. I then joined the Adaptive and Intelligent Robotics Lab (AIRL) as a Research Associate in Robot Learning and Fast Recovery, working with Professor Antoine Cully on DARPA's TRUSTLINE project, which is part of the Learning Introspective Control (LINC) program. TRUSTLINE develops machine learning-based introspection and monitoring technologies that enable robotic systems to detect and understand ongoing situations as they encounter uncertainty or unexpected events, while communicating these changes to human or AI operators to maintain confidence and operational continuity.
Beyond robot manipulation, my research extends to mixed-variable and combinatorial Bayesian optimization, with applications to autonomous antibody design and logic circuit optimization. I am passionate about pushing the boundaries of robot learning, developing methods that dramatically reduce data requirements and enable robots to be deployed effectively in everyday environments. If you'd like to discuss my research or just have a chat, feel free to reach out!