This defense addresses generalization under distribution shift with limited data. It introduces (1) diffusion-based inverse task inference that recovers a task embedding from a few demonstrations, enabling compositional generation without fine-tuning; and (2) bilinear transduction that converts out-of-support inputs into out-of-combination problems, yielding zero-shot extrapolation in robotics and property prediction.

This research explores why people form emotional bonds with social robots. Through forum analysis and a year-long self-study, it shows that humans transfer emotion to robots and experience reciprocal affect. The work proposes a new framework for understanding human–robot companionship as emotionally co-created, not purely technological.

Understanding how the brain controls behavior is key to studying neurological disease. This research introduces a high-speed robotic system that tracks mouse behavior in fine detail. By synchronizing precise behavioral data with brain activity recordings, it enables researchers to link specific neural regions to actions, improving insight into disorders like Parkinson’s and Alzheimer’s.

This research develops an AI-powered, multi-sensor drone system to detect butterfly landmines more safely and efficiently. By fusing sensor data into confidence-scored maps, it accelerates demining, reduces risk to operators, and supports civilian safety, land reuse, and humanitarian recovery in post-conflict regions.

This thesis introduces Armando, a low-cost soft robotic gripper with proprioceptive sensing using a single flexible capacitive sensor and neural-network decoding. Achieving 99% accuracy, Armando enables precise finger-position estimation for applications in prosthetics, assistive care, and disaster response, advancing accessible tactile robotics inspired by human touch.

Labour shortages leave millions of dollars of crops unharvested. This research develops touch-sensitive robots that navigate complex plants using force sensors rather than vision, reducing damage and improving fruit-reaching success by 66%. By learning from human movements, these robots could support sustainable agriculture and address critical workforce gaps.

This talk explains research that teaches legged robots how to walk reliably using machine learning, computer vision, advanced control theory, and Lyapunov-based safety guarantees. By improving robot stability on complex terrain, the work moves us closer to versatile, household multi-purpose robots capable of performing everyday chores safely and independently.

This project develops a 200-metre space reflector antenna using a modular “LEGO-like” assembly system. Designed for compact launch and robotic construction, it enables stronger, higher-quality interstellar communication. The work also models structural behaviour during assembly and could support building other large space structures, advancing deep-space exploration.