This project focuses on the development of a robotic planner that can learn to perform complex manipulation tasks, emphasizing the detection and application of precise forces and pressures during interaction with objects. One of the core challenges in robotic manipulation is enabling robots to execute delicate maneuvers, such as inserting a spatula under a pancake to flip it. This requires not only advanced control mechanisms but also the ability to generalize from limited data. As collecting large datasets to complete very specialized and niche tasks is often impractical in real-world scenarios, this research aims to explore data-efficient learning strategies for robotic planning and control.

I am participating in SuperUROP to leverage the experience I have gained in my previous research to work this problem. I am excited to apply my knowledge and experiences to formulate and propose my own solutions to interesting and challenging problems in robotics and planning. I aim to contribute to the field through the work I do in this coming year.