Control and Optimal Design of Wearable Robots
Summer school 2020
Universidad de los Andes
June 16-27, 2020. 8 AM - 12 PM
One of the biggest challenges in robotics is to design robotic systems that cooperate with humans. In the case of wearable robots, exoskeletons and robotic prosthetic legs are particularly challenging because the human is in the loop and poor mechanical or control design could have a significant impact on the health of the user. In this course, we will discuss the optimal design and control of these wearable robots using techniques from convex optimization and the experiences with the Open Source Leg. Some interesting questions we will solve in this class are: How to design elastic elements to minimize the energy consumption of the wearable robot? What is the role of convex optimization in these designs? How to mechanically communicate between the subject with amputation and the prosthetic leg?
First, let's build some motivation. Here are some of the central questions we will discuss in our class. (Curated videos to build our appetite)
How to synchronize a robotic knee and ankle with the intention of the user? Notice how the leg synchronizes with the user, especially during the first 17s of the video. One of the central questions of our class is: how to design a controller that communicates mechanically with the user?
Inspiring to see a bionic dancer! In terms of control of robotic prostheses, what do you think are the differences between amputation above the ankle (transtibial) or above the knee (transfemoral)?
Agonist-antagonist Myoneural Interfaces (AMI). How do you think biomechanisms should communicate with our body?
Myoelectric control and the machine state paradigm. More options to connect to our body.
How to design actuators that can interact with the environment? How can these ideas apply for the design of prosthetic legs?
What is convex optimization? Why is it important? How can we apply it to the design of robotic prosthetic legs?
How could technology impact our community and why does it matter?
Program
Pre-requisites
This class is self-contained and is designed for senior undergraduate or graduate students interested in robotics (e.g., Mechanical Engineering, Electrical and Electronics Engineering, Biomedical Engineering, and Computer Science.). The students should have: exposure to physics-based dynamic modeling, exposure to differential equations, strong background in linear algebra to understand the details of convex optimization, and Matlab programming. No previous experience in mechanical design, actuator design, or control is required.
Basically, read and understand this material before the first day of class: A Primer on Matrices, Stephen Boyd.
Instructors
Edgar Alberto Bolívar Nieto, Ph.D.
Edgar Bolívar is a Research Fellow at the Locomotion Control Systems Laboratory (LoCoLab) at the University of Michigan. The mission of the LoCoLab is to develop high-performance wearable control systems to enable mobility and improve the quality of life for persons with disabilities. Edgar is known for his work in the optimal design of series elastic actuators. He discovered that multiple objectives in the design of series elasticity, e.g., motor energy consumption and feasibility of torque-speed motor constraints could be formulated as robust-convex optimization programs. These contributions have been recognized with the Best Student Robotics Paper Award at the 2017 ASME Dynamic Systems and Control Conference. Edgar earned his Ph.D. and M.S. in Mechanical Engineering from The University of Texas at Dallas; and his B.S. in Mechatronics Engineering from the Universidad Nacional de Colombia.
Luis Felipe Giraldo Trujillo, Ph.D.
Luis Felipe received the PhD degree in Electrical and Computer Engineering at The Ohio State University, USA, in 2016. He is an Assistant Professor at Universidad de Los Andes. His current research interests include modeling and analysis of interconnected systems, applied machine learning, and humanitarian engineering.
Guest speakers (15- 20 min Q&As)
Robert Gregg, Ph.D.
Associate Professor
Director, Locomotor Control Systems Laboratory
Department of Electrical Engineering and Computer Science
Core Faculty, Robotics Institute
University of Michigan
Gray Thomas, Ph.D.
Research Fellow, University of Michigan
NASA Space Technology Fellow
Formerly at IHMC
Dario Villarreal, Ph.D.
Research and Innovation Manager
Toyota North America, USA
Department of Electrical Engineering
Southern Methodist University
Elliott Rouse, Ph.D.
Assistant Professor
Director, Neurobionics Lab
Department Mechanical Engineering
Core Faculty, Robotics Institute
University of Michigan
Google X
Class notes (Work in progress)
Convex Optimization
Convex Optimization_Reduced.pdfControl of Robotic Prosthetic Legs_Reduced.pdfControl of Robotic Legs
Actuation Technologies for Wearable Robots
Actuation Technologies for Robotics_Reduced.pdf