- Robot Learning for Manipulation
- Motion for Manipulation Tasks
- Learning and Classification
- Force and Tactile Sensors
- World Modeling
- Robot Programming by Demonstration
My research focuses on developing algorithms and representations to enable robots to learn versatile manipulation skills over time. The ability to learn skills and adapt manipulations to new situations will open up a wide range of new robot applications, including taking care of the elderly, maintaining parks and public places, and assisting in hazardous environments. I have developed methods for robots to learn about objects through physical interactions and improve their skills autonomously using reinforcement learning. I have also proposed representations for capturing various aspects of manipulations, e.g., contact states and motor primitives, to improve generalization between different scenarios and skills. The ultimate goal of my research is to develop a life-long learning framework for robots to acquire manipulation skills.
Before joining the CMU Robotics Institute in 2018, I was a postdoctoral researcher at the University of Southern California (USC). I received my Masters and Bachelors degrees in engineering from the University of Cambridge in 2008. From 2009 to 2011, I was a Ph.D. student at the Max Planck Institute for Intelligent Systems. In 2014, I defended my Ph.D. thesis at the Technische Universitaet Darmstadt and was a finalist for the 2015 Georges Giralt Ph.D. Award for the best robotics Ph.D. thesis in Europe.