TY - GEN
T1 - Predictive Modeling of Periodic Behavior for Human-Robot Symbiotic Walking
AU - Clark, Geoffrey
AU - Campbell, Joseph
AU - Rezayat Sorkhabadi, Seyed Mostafa
AU - Zhang, Wenlong
AU - Amor, Heni Ben
N1 - Funding Information:
1G. Clark, J. Campbell, and H. Ben Amor are with the School of Computing, Informatics, and Decision Systems Engineering, Arizona State University {gmclark1, jacampb1, hbenamor}@asu.edu 2S.M.R. Sorkhabadi and W. Zhang are with The Polytechnic School, Arizona State University, Mesa, AZ, 85212, USA. {srezayat, wenlong.zhang}@asu.edu This work was funded by the National Science Foundation under the Career Award grant number: FP00012258 Fig. 1. PIP is used to learn a realtime, closed-loop controller for symbiotic walking by analyzing and predicting sensor values, as well as control signals.
Publisher Copyright:
© 2020 IEEE.
PY - 2020/5
Y1 - 2020/5
N2 - We propose in this paper Periodic Interaction Primitives - a probabilistic framework that can be used to learn compact models of periodic behavior. Our approach extends existing formulations of Interaction Primitives to periodic movement regimes, i.e., walking. We show that this model is particularly well-suited for learning data-driven, customized models of human walking, which can then be used for generating predictions over future states or for inferring latent, biomechanical variables. We also demonstrate how the same framework can be used to learn controllers for a robotic prosthesis using an imitation learning approach. Results in experiments with human participants indicate that Periodic Interaction Primitives efficiently generate predictions and ankle angle control signals for a robotic prosthetic ankle, with MAE of 2.21° in 0.0008s per inference. Performance degrades gracefully in the presence of noise or sensor fall outs. Compared to alternatives, this algorithm functions 20 times faster and performed 4.5 times more accurately on test subjects.
AB - We propose in this paper Periodic Interaction Primitives - a probabilistic framework that can be used to learn compact models of periodic behavior. Our approach extends existing formulations of Interaction Primitives to periodic movement regimes, i.e., walking. We show that this model is particularly well-suited for learning data-driven, customized models of human walking, which can then be used for generating predictions over future states or for inferring latent, biomechanical variables. We also demonstrate how the same framework can be used to learn controllers for a robotic prosthesis using an imitation learning approach. Results in experiments with human participants indicate that Periodic Interaction Primitives efficiently generate predictions and ankle angle control signals for a robotic prosthetic ankle, with MAE of 2.21° in 0.0008s per inference. Performance degrades gracefully in the presence of noise or sensor fall outs. Compared to alternatives, this algorithm functions 20 times faster and performed 4.5 times more accurately on test subjects.
UR - http://www.scopus.com/inward/record.url?scp=85092718871&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85092718871&partnerID=8YFLogxK
U2 - 10.1109/ICRA40945.2020.9196676
DO - 10.1109/ICRA40945.2020.9196676
M3 - Conference contribution
AN - SCOPUS:85092718871
T3 - Proceedings - IEEE International Conference on Robotics and Automation
SP - 7599
EP - 7605
BT - 2020 IEEE International Conference on Robotics and Automation, ICRA 2020
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2020 IEEE International Conference on Robotics and Automation, ICRA 2020
Y2 - 31 May 2020 through 31 August 2020
ER -