Toward Human Motion Digital Twin: A Motion Capture System for Human-Centric Applications

作者：Huiying Zhou, Longqiang Wang, Gaoyang Pang, Huimin Shen, Baicun Wang, Haiteng Wu, Geng Yang · 发表于：IEEE Transactions on Automation Science and Engineering · 年份：2024 · DOI：10.1109/tase.2024.3363169 · 被引用次数：29 · 研究领域：Digital Transformation in Industry、Hand Gesture Recognition Systems

Following the rule of human-centricity, Human Motion Digital Twin (HMDT) attempts to apply human motion data to ensure the development and well-being of human beings. Particularly, perception and estimation of human motion play fundamental roles in realizing HMDT. This work proposes an inertial motion capture system for human motion digital twin (InMoDT). The designed motion capture device is made up of a hub node and inertial measurement units attached to the human body. The proposed algorithm framework supported by sensor fusion and pose calibration algorithms, enables to acquire orientations of sensors and body segments. With the deployment of algorithms, InMoDT achieves an average root mean square error of 4.7$^{\circ}$in estimating orientations when compared with an optical motion capture system. Experimental results show a great correlation ($92.5\%$) and agreement ($97.8\%$) between InMoDT and the optical system. The abilities of InMoDT are spotted in terms of human-centric applications based on the integration of human, cyber system, and physical system, such as motion monitoring and estimation, and human-robot teleoperation.Note to Practitioners—This paper is motivated by the problem of inadequate attention on humans in Cyber-Physical System (CPS) while the roles of operators have a significant effect on industry. With the popular applications of digital twins in CPS, HMDT is expected to monitor, analyze, and assess motion data for facilitating the Human-Cyber-Physic...