Learning with Contacts

Learning through contact is one of our future directions. As robots venture from traditional static environments to versatile ones, they need to autonomously acquire new skills. This project aims to explore the capabilities enabled through force or tactile sensing. By leveraging these sensing technologies, robots can improve their ability to recognize and manipulate objects, perform complex tasks, and adapt to dynamic environments. This approach involves utilizing various learning mechanisms to enhance robots’ adaptability and functionality, ultimately making them more capable and versatile in a range of applications.

University of Toronto Institute for Aerospace Studies