1 Introduction
Traditional industrial systems, where physical systems, i.e., robots and machines, replaced manual labor mostly, are often known as human-physical system (HPS) [1]. Cyber system, a major outcome of the digitalization revolution, is increasingly adopted in HPS and continuously evolving with the development of advanced manufacturing technologies, leading to the human-cyber-physical system (HCPS) [2]. HCPS fuses human, social networks, physical processes, and cybernetics as an integrated system, characterized by intelligence interaction, diversified integration, and grand system. HCPS connects human, virtual and physical worlds by incorporating intrinsic and abstractive knowledge learned from a variety of manufacturing activities. With the development of advanced sensing, actuation, embedded computing, and artificial intelligence (AI), HCPS will not only create a harmonious human-machine-intelligence collaboration paradigm but also become applicable in a wider range and more sophisticated industrial and health care use cases.
The growing advanced manufacturing technologies have considerably morphed the role and the responsibilities of human in industrial system. In the traditional cyber-physical system (CPS), most research focuses on developing more autonomy, creativity and responsibility of intelligent machines, which inadvertently overlook human’s importance in critical decision making. One of the major objectives in HCPS is to account for the human in the loop, i.e., shifting from technology-driven approach to a human-centric approach. There are many challenges to realize the HCPS in practice. For example, multi-source context perceptions in the dynamic environment and real-time control of heterogeneous equipment limit the situation awareness of machines thus hindering implementation of HCPS. To tackle the above challenges, the human digital twin (HDT) is introduced to perceive the current context of HCPS and recommend solutions for the human and physical systems.
HDT highlights the importance of human in integrating the physical world and virtual world in HCPS [3]. HDT includes physical representations and virtual models of humans to accurately track and reflect the human motion, perception and manipulation activities and capabilities and to address the challenges in the human-centric manufacturing. Its interconnection and integration among each element regulate human-machine alignment based on human’s proactivity, depict the human-machine interaction and operation performance with the physical and digital representations.
HCPS envisages a system that couples human and physical system to a virtual world. In the context of the HCPS, digitalization of human and physical system helps human to focus on high-value-added tasks and use machines and robots to amplify their impacts on production. Furthermore, increasingly intelligent physical system can free up human capitals and creativity, which makes humans easier to achieve their goal. Here, we propose a HDT-driven HCPS to address such needs to enable human and machines to evolve together in a harmonized way. Moreover, in some special areas, such as the underwater maintenance and space exploration scenarios, the tasks occur in unreachable locations at which, due to accessibility, flexibility, and other technical limitations, the physical representations and virtual models of human are of necessity [2]. To our best knowledge, no referenced HCPS model is structured to appeal to human-centric demands, handle the unstructured working environment and grasp the emerging opportunity that has arisen from the breakthrough in advanced manufacturing technologies. Though studies have been reported the potential application scenarios of HDT in different systems, a unified understanding and summarization of HDT-driven HCPS’s framework and development methods remains to be defined and clarified. The contribution of this paper is to a comprehensive framework of HDT-driven HCPS to address the following issues (1) - distributing skill-based tasks and workload between human and physical system accounting for the variability in manufacturing systems (2) realizing human-physical system interaction at the system level, providing and enhancing transparency in human and physical systems, and enhancing HCPS’s intelligence regarding analytical assessment, predictive diagnosis, and performance optimization. (3) interconnecting human and the physical system closely and thoroughly by human-machine consistency, and, determining the appropriate level of human-physical system collaboration. In this study, we have discussed the framework, enabling technologies, and application examples of HDT-driven HCPS to fulfill the research gap.