首页|New Findings from University of Alberta Describe Advances in Androids (Uncertainty-aware Safe Adaptable Motion Planning of Lower-limb Exoskeletons Using Random Forest Regression)
New Findings from University of Alberta Describe Advances in Androids (Uncertainty-aware Safe Adaptable Motion Planning of Lower-limb Exoskeletons Using Random Forest Regression)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Elsevier
Researchers detail new data in Robotics - Androids. According to news reporting from Edmonton, Canada, by NewsRx journalists, research stated, “Human safety and data security are two of the main concerns that have limited the utilization of deep learning based techniques in medical robotic applications. Such concerns are amplified by uncertainty in the deep learning run-time predictions.” Financial supporters for this research include Canada Foundation for Innovation, Government of Alberta, Natural Sciences and Engineering Research Council of Canada (NSERC), Canadian Institutes of Health Research (CIHR), Alberta Economic Development, Trade and Tourism Ministry’s grant. The news correspondents obtained a quote from the research from the University of Alberta, “In this paper, we propose a novel framework for incorporating uncertainty analysis that is fast enough (updates in 20 Hz) to be used in the control loop of a medical robot and that considers both the training and testing phases of the deep learning algorithm. As a case study focusing on the use of a lower-limb exoskeleton to assist the walking of people with disability, we learn the passive human- exoskeleton system’s dynamics using Random Forest Regression (RFR) and quantify the uncertainty level of its prediction. Whereas prior art fed the estimated human-robot interaction torque values to the adaptable Central Pattern Generators (CPGs) to refine the gait trajectories, our contribution is to leverage the knowledge of the predictions’ uncertainty levels to ensure safety in human-robot interaction. Our proposed framework for uncertaintyaware control of medical robots finds the similarities of labels and predictions in the training set using Kullback-Leibler (KL) divergence, while in the test phase, it detects out-of-distribution (OOD) data using Mahalanobis distance between test feature and training distribution. As compared to state-of-the-art methods, the proposed method is real-time and addresses the issue of uncertainty in the decisions of the robot controller. We have tested the proposed method on ExoH3 (Tehnaid S.L.) lower-limb exoskeleton. The experiments were conducted to evaluate the performance of the uncertainty analysis technique. The results demonstrate that our proposed uncertainty analysis technique can detect OOD features resulting in unsafe motion planning.”
EdmontonCanadaNorth and Central AmericaAndroidsEmerging TechnologiesHuman-Robot InteractionMachine LearningRobotRoboticsUniversity of Alberta
NETL
NSTL
Elsevier
