This $522,592 Project Grant awarded by the National Science Foundation (NSF) Engineering program (CFDA 47.041) aims to develop a data-driven autonomy framework that enables non-technical users to safely and effectively teach robots. The research has three interconnected thrusts:
-
Creating a novel framework for data-driven control of complex mechanical systems, focusing on stability and safety. This includes a Koopman control factorization theory to enable efficient learning and control synthesis.
-
Introducing methods for teaching robots through user demonstration, using stable feedback-feedforward controllers to replicate taught behaviors while safeguarding against hardware limitations.
-
Addressing long-term adaptability to changing environmental conditions through online system identification and active learning.
The theoretical advancements will be validated through practical outdoor testing to demonstrate robustness in real-world environments. This research has the potential to transform sectors like agriculture by fostering a more resilient and adaptable supply chain.
Generated 1/28/25, 6:55 AM