Electrical and Computer Engineering Associate Professor Joohyung Kim has led the research collaboration between the University of Illinois Urbana-Champaign and Toyota Research Institute (TRI) with the goal of amplifying the human experience. For the last two years, Kim has researched multimodal sensing methods for robotic manipulation by exploring soft/hard materials, fabrication methods, and design optimization to develop useful sensing methods for robotic tasks to help people in their daily lives at home. Kim recently gave some insight into his involvement with TRI over the last couple of years.
What is the research that you do in collaboration with TRI?
JK: In 2021, I started a project called “Superhuman Multimodal Sensing for Manipulation” funded by TRI. The goal of this project was to combine multimodal sensors with fabrication techniques for robotic soft structures to enhance traditional vision-based perception for manipulation tasks.
What is the impact of your research at TRI?
JK: For this project, I have been exploring soft materials and fabrication methods to design robotic “grippers” and 3D printed skin sensors for robotic systems. We also provided fabrication support with thermoplastic polyurethane (TPU) airbags for robotic arms and membrane sensors for soft bubble grippers that can recognize objects by touch.
Since your involvement with TRI began, what has been your biggest accomplishment?
JK: We are proud to have developed several methods for fabricating artificial, soft skins with TPU that have improved the grip of robotic arms. The “Lip-Inspired Passive Jamming Gripper” was the first gripper we developed using these advancements and is inspired by canine oral structures. We have already made progress on a second generation of the gripper that has an improved design and a camera that enhances grasping performance. We recently submitted a paper about this work!
What do you hope comes out of your work with TRI in the future?
JK: Through this project, I hope we can expand the sensory space for tasks in cluttered and unpredictable environments such as a home kitchen by going beyond computer vision. Soft, adaptable touch and force sensing are vital to enabling more robust perception schemes for dexterous, tactile-driven manipulation.