Please join us for a lecture from our Faculty Candidate: Yuri Gloumakov – Post Doc at University of California, Berkeley
ITE 336 on Monday March 25, 1pm
Title:
Learning from human performance for semi-autonomous robot control
Abstract:
The loss of upper-limb function continues to have a profound impact on one's independence and overall quality of life, highlighting the critical need for innovative solutions in the realm of assistive devices and rehabilitation. Fully restored function would include nearly effortless end-effector positioning, grasping, and an ability to perform in-hand manipulation, and yet advancement in upper-limb prostheses have largely focused only on grasping. In this talk, I address both positioning and in-hand manipulation using model-free and model-based approaches. In this talk I show how data-driven tools are leveraged to distill human arm movements into distinct actionable categories. These are then validated in a virtual reality experiment, where I demonstrate that this novel approach can improve the operation of an n-DOF upper-limb device when tested against conventional methods. I will conclude by presenting a model-based approach that can be used with geometrically featured tactile sensors to perform human-like dynamic in-hand manipulation. These findings help us better understand and emulate upper-limb function and bring us a step closer to embodied assistive devices.
Join from the meeting link
https://uconn-cmr.webex.com/uconn-cmr/j.php?MTID=m39c4465c205a12d389c0c4ed56c0f243
For more information, contact: Brandy Ciraldo at brandy.ciraldo@uconn.edu