In the last few decades, society has sought robotic solutions to human mobility problems. From assisting weakened limbs with robotic exoskeletons to filling in for missing limbs with robotic prostheses, these technologies have uncovered a number of challenges related to interacting physically with a human body.
In this talk, I will discuss how the Robotics & Technology for Human Health and Mobility (RTHM) lab is addressing the challenges of intent recognition and improved motor learning for lower limb assistive robots.
The ‘intent recognition’ problem refers to the issue of determining what movements the human user wishes to accomplish so that the assistive robot can implement an appropriate controller to assist with that activity. While previous work has sought to glean this information from measurements of the human body, such as electromyography and electroencephalography, we are instead working to incorporate computer vision of the user’s external environment. The idea is that the objects in the environment, their associated uses, and the user’s gestures near these objects hold a wealth of information related to the user’s intended actions.
Intent recognition is most important for assistive robots that are meant to be used out in the real, cluttered, and unstructured world of users’ day-to-day lives. In contrast, we are also interested in how assistive robots can be used in a structured physical therapy environment. While intent recognition is less important in this context, these robots face the additional challenge of aiming to affect how humans learn motor skills. In other words, we aim for the user to gain a skill by interacting with the robot such that the skill is retained even once the robot is removed.
In this part of the talk, I will discuss two projects in the RTHM lab – 1) quantifying the optimal challenge point for simple game-like tasks and 2) characterizing how humans learn complex balance-challenging tasks. Both projects aim to improve rehabilitation outcomes for individuals undertaking robot-assisted physical therapy.
While my talk will not address my professional journey following my graduation from Notre Dame, I am happy to answer questions about my experience as a postdoc and early-career faculty after leaving South Bend.
Originally from Simpsonville, SC, Prof. Taylor Higgins earned her B.S. in Mechanical Engineering from Clemson University. In 2022, she graduated from the University of Notre Dame with her Ph.D. in mechanical engineering advised by professors Jim Schmiedeler and Pat Wensing. Her dissertation work specifically aimed to identify human gait speed intentions for lower-limb exoskeletons.
After Notre Dame, she took a brief postdoc position at UT Austin with Dr. Ann Majewicz Fey working on optimal haptic guidance strategies for surgical robotics. In 2023, she started as an assistant professor at Florida State University in Tallahassee where she continues to work to increase the fluency of a broad spectrum of human/robot interactions as the director of the Robotics & Technology for Human Health and Mobility (RTHM) lab.