Home About the Journal Latest Work Current Issue Archive Special Issues Editorial Board

TABLE OF CONTENTS

2019,  1 (6):   558 - 579

Published Date:2019-12-20 DOI: 10.1016/j.vrih.2019.09.005

Abstract

Virtual reality (VR) has been widely used in various manufacturing industries, and VR-based virtual manufacturing has received significant attention in the current intelligent manufacturing era. Digital human models (DHMs) are essential for virtual manufacturing applications. Additionally, researching new applications of DHMs has developed into an important academic research field. This paper aims to identify the applications and research trends of DHMs in the manufacturing industry and to provide a reference for the continued development of virtual manufacturing and DHMs. We selected a total of 49 related articles from a large number of articles published between 2014 and 2019. The applications of DHMs in the manufacturing industry are analyzed from different perspectives and various relevant technical limitations are discussed. The results indicate that the applications of DHMs differ significantly between different types of fields. The automotive industry is the main application field for DHMs, and assembly/maintenance simulations and evaluations are the main application types. Additionally, there are still some limitations in the establishment of virtual environments, motion control, and DHM evaluation that should be addressed. Finally, research trends in the application of DHMs are illustrated and discussed, including the planning and assessment of human-robot collaboration systems, the combination of DHMs and augmented reality, and improved motion planning for DHMs. In summary, the application of DHMs can improve the realism and effectiveness of virtual manufacturing, and DHMs will be more widely and deeply studied and applied in various manufacturing industries in the near future.

Content

1 Introduction
Virtual reality (VR) systems generate virtual environments using computer modeling and simulation technology, allowing users to have an immersive experience by interacting with virtual environments, which can provide an effective method for people to learn about the world[1,2]. VR integrates multimedia, sensor, display, human-machine interaction, ergonomics, simulation, computer graphics, and artificial intelligence technologies to expand human perception. VR technology has been widely used in education, healthcare, entertainment, culture, sports, engineering, the military, and other sectors[3,4,5]. VR has existed in some form for more than half a century, but has only recently developed into a practical tool for manufacturing industries. As a result, VR-based virtual manufacturing has received significant research attention as a novel technology. Based on the rapid development of manufacturing technology and constant changes in customer demands, modern manufacturing enterprises face significant pressure from global competition. Virtual manufacturing is becoming an important technology for coping with competitive pressures and facilitating the transformation and upgrading of enterprises[6,7]. Virtual manufacturing can play an important role in every phase of the product lifecycle. For example, in the product design phase, virtual manufacturing can analyze future manufacturing processes and other activities in the product lifecycle to ensure the optimization of product design quality and maximization of production efficiency[8,9].
Currently, automation equipment, such as industrial robots and computer numerical control machines, is widely used in manufacturing industries, but there are many activities that require manual operations [10]. For example, automation equipment is currently unable to perform various complex and flexible tasks. Additionally, manual operations are still dominant in most small- and medium-sized enterprises. Therefore, human operators still play a major role in many manufacturing industries[11]. In most cases, it is necessary to consider human factors/ergonomics (HFE) issues during the product design process, which can introduce many benefits. First, HFE issues are important for product design[12,13]. Many problems that may arise in the future can be identified in the design stage by considering HFE issues, allowing a design scheme to be improved in its infancy to mitigate hidden dangers[14]. For example, the relationships between environments and operators should be considered in the design of production lines, workstations, and automotive cabins[15,16]. Second, HFE issues are important for process planning in manufacturing industries[17]. For example, to avoid the risk of musculoskeletal disorders (MSDs), the work postures and processes of operators should be analyzed and optimized in product assembly and maintenance planning[18]. Therefore, the consideration of HFE issues and adoption of digital human models (DHMs) in virtual manufacturing have become widespread, which should improve the realism and effectiveness of virtual manufacturing technologies.
Complete DHMs should represent all functions of a real human, including anatomy, motion, dynamics, biomechanics, psychology, and physiology. However, under current technical conditions, it is difficult and unnecessary to establish complete DHMs. For various application scenarios, DHMs are constructed by simplifying real human models according to different needs[19,20]. In general, DHMs can be divided into two categories: cognitive DHMs and physical DHMs[12,21]. Cognitive DHMs focus on the realization of human-machine interaction based on emotions, languages, and facial expressions, and are mainly applied in the journalism and social psychology fields. Physical DHMs focus on the functions of the body structures of real humans and can be subdivided into two types: physiology-related DHMs and posture- and motion-related DHMs[22]. Physiology-related DHMs are constructed for research in the fields of biomedicine, radiology, automotive safety, etc. Posture- and motion-related DHMs are constructed for research in the field of ergonomics and focus on working postures and comfort. In manufacturing industries, posture- and motion-related DHMs are the dominant DHM form[23]. The objective of the paper is to identify the applications and research trends of DHMs in manufacturing industries. Therefore, posture- and motion-related DHMs are the main focus in this paper.
Overall, this paper is organized as follows. Section 1 introduced the background, scope, and purpose of our study. Section 2 describes the literature search strategy utilized in this study. Section 3 presents an overview of the applications of DHMs in manufacturing industries. Section 4 illustrates and discusses various research trends and summarizes the outcomes of this study.
2 Search strategy
To identify the applications and research trends of DHMs in manufacturing industries, a systematic and complete literature search was necessary. Based on the search strategies proposed by MacDonald[24] and Santos[25], we divided the literature search process into three steps: planning, searching, and assessing. Specific methods are discussed below.
2.1 Planning
The tasks for the planning phase included defining research questions, setting a timescale for searching, and selecting databases. Research questions were formulated to address the objective of our study, which was to identify the applications and research trends of DHMs in manufacturing industries. Research questions can aid researchers in analyzing the relevance of retrieved documents. The following research questions are considered in this paper.
Q1: What are the current applications of DHMs in manufacturing industries?
Q2: What application environments and development tools have been used for the application of DHMs?
Q3: What are the key technologies for the application of DHMs?
Q4: What are the research trends for DHMs in manufacturing industries?
In recent years, the rapid development of DHMs in manufacturing industries has led to the emergence of a large amount of published research. Therefore, considering space limitations, we limited our search to the past five years from January of 2014 to May of 2019. This timescale will enable readers to understand the state-of-the-art of DHM applications and while limiting search work.
To ensure the quality and integrity of retrieved documents, the following six databases, which have been widely recognized and used by researchers around the world, were selected for this study[26,27,28,29].
• IEEE Xplore (www.ieeexplore.com)
• ScienceDirect (www.sciencedirect.com)
• Scopus (www.scopus.com)
• Google Scholar (www.scholar.google.co.uk)
• Web of Science (www.isiknowledge.com)
• Engineering Village (www.engineeringvillage.com)
2.2 Searching
In the search phase, the six databases above were used separately to retrieve documents. Because manufacturing industries touch many application fields, the choice of search terms is very important. Suboptimal search terms could lead to the omission of many useful documents. Based on the research objectives and questions, we expanded the search scope for this study by using “digital human” as a search term. Additionally, because the term “virtual human” has the same meaning as “digital human” in many studies, we utilized the search string (“digital human” OR “virtual human”). The search settings and results for different databases are listed in Table 1. Some documents retrieved from the databases were duplicates. After removing duplicate documents, 1412 documents remained.
Search settings and results
Database Search Fields Quantity
IEEE Xplore Metadata Only 164
ScienceDirect Title-Abs-Key 145
Scopus Title-Abs-Key 1209
Google Scholar Title 701
Web of Science Title-Abs-Key 1107
Engineering Village Subject-Title-Abs 694
2.3 Assessment
The 1412 retrieved documents required further assessment. The goal of the assessment phase was to identify the most relevant documents and reduce the total number of documents based on analysis and filtering. Exclusion criteria were defined to conduct the first screening of documents, similar Riccardo’s study[29].
The following exclusion criteria were defined for this study.
(EC1) The document is not in English.
(EC2) The document is not a journal article, confe-rence article, or review article.
(EC3) The document is not related to a manufacturing industry.
(EC4) The document is not related to posture- and motion-related DHMs.
(EC5) The document focuses directly on DHMs, rather than their applications in manufacturing industries.
The first screening process and its results are illustrated in Figure 1. A total of 53 documents remained following the first screening process. Next, the quality of each document was evaluated to identify the most appropriate documents. Santos presented five quality criteria for evaluating documents[25]. In addition to these criteria, we also considered document type as a quality criterion, as shown in Table 2. Journal articles are assigned one point, while conference and review article are assigned 0.5 points and 0.75 points, respectively. The second screening process remo-ved any documents with a quality score less than two. The evaluation results are listed in Table 3. Four documents were removed in the second screening process, leaving 49 documents. Among the remaining documents, there are 21 journal articles, 25 conference articles, and three review articles[30,31,32].
Quality Criteria
ID Description Points
QC1 The document is described clearly? 0/0.5/1
QC2 The methods or techniques are presented clearly? 0/0.5/1
QC3 The constructed virtual environment is described clearly? 0/0.5/1
QC4 The study is evaluated and validated clearly? 0/0.5/1
QC5 The results of case studies are provided in detail? 0/0.5/1
QC6 The document type is a journal/conference/review article? 0.5/0.75/1
Evaluation results
Score Quality Quantity
[5,6] Good 11
[2,5) Average 38
[0,2) Poor 4
3 Applications of DHMs in manufacturing industries
In this section, an overview of the applications of DHMs in manufacturing industries is presented by answering research questions Q1 to Q3 based on six factors: application field and application type (Q1), application environment and development tools (Q2), and motion control and HFE evaluation method (Q3).
The 21 selected journal articles and 25 selected conference articles are considered to represent recent studies. These articles are analyzed and discussed in this section. Additionally, the statistical charts presented in this section were derived from these 46 documents. The three review articles are not included in these statistics, but their results will be considered and utilized in this paper.
3.1 Application field
Similar to Riccardo’s method[29], the application field of a paper was defined as the industry and/or technological environment in which the application of DHMs was carried out. According to the content and statistical analysis results of the 46 articles, their application fields were categorized and the categories were analyzed and selected for further analysis. For example, there was a very small number of articles from the oil industry and nuclear industry. There is no need to define each application field as a category, which would result in a very high number of categories. Therefore, in this study, the oil industry and nuclear industry were combined into the energy industry. The application fields of DHMs can be divided into the following five categories: (1) Automotive industry; (2) Industrial plants; (3) Aerospace industry; (4) Military industry; (5) Energy industry.
These five categories are not abstracted on the same level and they are not necessarily mutually exclusive. In general, an application field is classified according to the industry in which the corresponding technology is applied. However, some articles did not specify an industry, such as [33] and [34], and only indicated that the manufacturing environment was an industrial plant. According to the representation and emphasis of these articles, it is feasible to classify application fields based on different levels of abstraction.
As shown in Figure 2, the automotive industry, industrial plants, and aerospace industry are the main application fields of DHMs. This result is consistent with those presented by Masaaki[30] and Bures[36]. The automotive industry is one of the most important, representative, and mature manufacturing industries. Cutting-edge technologies are often first applied and promoted in the automotive industry. Therefore, the automotive industry has a clear lead in the application of DHMs[36]. For industrial plants, the optimization of workplaces and production efficiency has always been the main research focus and related studies typically consider detailed HFE issues. With the application of virtual manufacturing in industrial plants, DHMs have been widely used[33,34]. The aerospace industry is a rapidly developing industry that plays a vital role in national security and economics, and it has received significant attention. Furthermore, in the aerospace industry, manual operations are dominant. Currently, DHMs are an important tool for designers who wish to analyze manual operations, so DHMs have also been widely used in this industry[37,38]. In contrast, the application of DHMs is relatively sparse in the other three fields. For the energy and military industries[39,40,41], based on their respective industrial characteristics, the popularity of virtual manufacturing technology is not widespread, meaning applications of DHMs are limited. However, with continued development and promotion of virtual manufacturing and DHMs, the application of DHMs in the energy and military industries should gradually increase[30].
3.2 Application type
Application types are defined by the specific industrial tasks to which DHMs are applied. Compared to application fields, application types can provide a deeper description of DHM applications. After reading the 46 selected articles in detail, their types were divided into the following six categories: (1) Assembly/maintenance simulation and evaluation; (2) Automotive interior assessment; (3) Workplace design and optimization; (4) Human-robot collaboration simulation; (5) Human behavior study; (6) Education/training.
The proportion of each application type is presented in Figure 3. The assembly/maintenance simulation and evaluation category were identified in 41.3% (19) of the articles. For mechanical products, assembly and maintenance are two important issues throughout the product lifecycle[43,44]. Virtual assembly (VA) and virtual maintenance (VM) have largely replaced physical prototyping and become the main means of design and planning for assembly and maintenance procedures. Some studies on VA and VM only focus on products without considering HFE issues, which can lead to operations that are difficult or impossible to complete[45,46,47]. Therefore, it is necessary to consider DHMs for assembly and maintenance simulation. First, DHMs are used to perform assembly/maintenance operations in a virtual environment according to a proposed design scheme. Next, the simulation process is evaluated, and problems related to assembly and maintenance can be identified in the design stage. Therefore, research on DHMs for assembly/maintenance simulation and evaluation focuses on two main aspects: improving the accuracy and efficiency of DHM movements[38,41,44,48,49,50,51,52,53,54] and evaluating the assembly/maintenance process more accurately[37,55,56,57,58,59,60,61,62]. For the former aspect, many scholars have studied motion control methods for DHMs, which will be outlined in Section 3.5. For the latter aspect, in addition to the common factors of HFE evaluation, other factors have also been considered. As shown in Figure 4a, Louison et al.[55] demonstrated that force feedback is necessary for interactions between real humans and DHMs in a confined environment, meaning the influence of tactile sensation must be considered in simulations and HFE evaluations. As shown in Figure 4b, He et al.[57] considered the influence of lighting and studied maintenance simulations and HFE evaluations of DHMs under different lighting conditions. As shown in Figure 4c, Geng et al.[62] considered the influence of maintenance time and developed a time compensation model to predict the required durations of real operations based on simulations in a virtual environment.
The automotive interior assessment category was identified in 17.4% (8) of the articles. Automotive interior systems are an important part of automobile design that directly determine the safety and comfort of occupants[63,64]. The efficiency and quality of automotive interior design have been improved significantly by applying virtual manufacturing and DHMs. The application of DHMs to automotive interior assessment includes two main types of assessment: cabin assessment[63,64,65,66,67,68] and seat assessment[69,70]. The assessment process typically involves three steps: (1) Setting up different interior configurations, such as driver’s seat height and control button position; (2) Utilizing various percentiles of human physical dimensions (i.e., 5th and 95th) and adjusting the postures of DHMs; and (3) complete HFE evaluations. Through the evaluation and comparison of different interior configurations, the optimal interior configuration can be identified. In most cases, DHMs only need to maintain static sitting postures or perform simple movements, as shown in Figure 5a. Additionally, as shown in Figure 5b, Wan et al.[65] studied the automobile ingress/egress process. They generated human body swept volumes based on the ingress/egress motions of DHMs and assessed the design of automotive cabins by evaluating interference between the swept volumes and automotive cabins.
The workplace design and optimization category were identified in 17.4%(8) of the articles. The workplace is an important part of manufacturing systems, and workers are typically required to work long hours in their workplace. Unreasonable workplace designs lead to low productivity and increase the risk of MSDs[71,72,73]. Typically, once the workplaces for a manufacturing system are established, adjustment costs are large. Therefore, detailed workplace design is important and virtual manufacturing technology has become the main method for this type of design task. The layout of a workplace can affect workers‘ visibility, reachability, comfort, etc. To design a reasonable workspace, different design div should be validated and compared based on simulations before a workspace is constructed. DHMs are used to complete the defined operations in simulations. Based on the HFE evaluation of DHMs, the rationality of a workplace layout can be determined[71,72]. Additionally, the same approach can also be used to improve existing workplace layouts[74,75,76].
The human-robot collaboration simulation category was identified in 10.9%(5) of the articles. Human-robot collaboration combines the repetitive performance of robots with the unique skills of human beings to realize novel production modes with high levels of flexibility and automation[42]. Before actual production begins, it is necessary to verify and evaluate the safety, collaboration, efficiency, and rationality of human-robot collaboration systems[35,77]. However, there are certain risks associated with using real robots for verification. Additionally, debugging real robots is time consuming and complex. Therefore, simulation methods are a viable alternative that can allow verification work to be completed in a virtual environment. During simulations, DHMs are used to complete interactions with virtual collaborative robots according to the target manufacturing process. By analyzing the interaction process, the performance of human-robot collaboration systems can be evaluated. Human-robot collaboration simulations include two main types: robots assisting workers to complete operations, where robots and workers can be treated as a single unit[35,42,78]; and robots and workers completing operations independently of each other[77,79]. For the former type, as shown in Figures 6a to 6d, Maurice et al.[77] studied the stress situation in each joint of a DHM under different configuration schemes. These stresses can be used to identify optimal configurations. For the latter type, as shown in Figures 6e to 6g, Castro et al.[42] evaluated different configuration schemes using a rapid upper limb assessment (RULA) spreadsheet and identified optimal configurations by comparing RULA scores. Additionally, they also considered the influence of human body dimensions in their study.
The human behavior study category was identified in 6.5%(3) of the articles. In studies on human behavior, the construction of experimental environments may be complex, costly, or dangerous. Currently, for simulating human behavior in manufacturing processes, some scholars have begun to utilize VR technology to construct high-fidelity virtual environments for human behavior studies[33,40,80]. The behavior of real humans in physical environments can be predicted based on the performance of DHMs in virtual environments. As shown in Figure 7, Gürerk et al.[33] studied peer effects in industrial plant production. In their experiments, a real human subject observed various productive virtual peers and altered his own working rate correspondingly. Additionally, Li et al.[40] studied the unsafe behavior of DHMs in virtual nuclear decommissioning operations. Generally, the accuracy of such VR-based human behavior studies largely relies on the authenticity of virtual environments and means of human-computer interaction.
The education/training category was identified in 6.5%(3) of the articles. Traditionally, in manufac-turing industries, education and training to provide skills and safety instructions are delivered through various documents and videos[81]. With the application of VR and DHMs, trainees can observe and interact with DHMs in a virtual environment, which increases learning interest and efficiency significantly[82,83]. Furthermore, in many cases, the cost of education and training can be reduced.
3.3 Application environments
Typically, according to the presentation type of a virtual environment, VR is divided into immersive VR and non-immersive VR. Immersive VR mainly presents virtual environments using cave automatic virtual environments (CAVE) or head-mounted displays (HMD), which can provide users with a strong and immersive feeling of presence[33]. In contrast, non-immersive VR presents virtual environments using desktop display systems, which cannot provide a high level of immersion[39]. Additionally, augmented reality (AR) technology can also be used to provide virtual information or scenes[63]. Therefore, in manufacturing industries, the application environments for DHMs can be constructed according to the methods described above. Typical applications of these types of environments in manufacturing industries are presented in Figure 8.
According to the classification outlined above, the types of application environments in the 46 selected articles were counted. The results are presented in Figure 9. As shown in Figure 9, desktop-based application environments of DHMs are the most widely used type in current practical applications, accounting for 63.0%(29) of the selected articles. This is mainly because of the popularization of desktop display systems and corresponding commercial software that provide DHM functionality. However, the application of CAVE and HMD systems is relatively uncommon, which may be related to two different factors. First, the establishment and operating costs of CAVE and HMD systems are relatively high[33,72], which limits their application. Second, the movements of DHMs are driven by real humans, meaning the workload for manufacturing process simulation is very large[48]. Furthermore, the debugging of related equipment in CAVE and HMD systems, such as the calibration of motion capture systems, is time consuming. Therefore, manufacturing enterprises do not prefer applying CAVE and HMD systems to practical applications. Additionally, some researchers have established both immersive and non-immersive systems (10.9%(5) of the articles) for two different objectives[41,44,52,72,75]: comparing the two types of application environments above and comb-ining them for DHM simulations. To reduce the workl-oad of modeling and processing virtual environ-ments and models, AR technology can be applied (4.3%(2) of the articles)[63,65]. For example, the authors of [63] combined DHMs with physical environments to achieve the goal of automotive interior assessment, as shown in Figure 8c.
3.4 Development tools
In this paper, the development tools for DHM applications are introduced from two perspectives: hardware and software[29,31]. The hardware and software used in the 46 articles selected articles were analyzed. Some results are listed in Table 4. For hardware, ubiquitous devices, such as computers, keyboard, and mouse, are not listed.
Development tools
Article Application type Application environment Hardware Software
[33] Human behavior study CAVE aixCAVE/Kinect/3D glasses/Motion capture system 3Ds Max/SmartBody
[35] Human-robot collaboration simulation HMD eMagin Z800 3DVisor HMD/Kinect Unity 3D/Evolver 3D digital avatar/3Ds Max
[37] Assembly/maintenance simulation and evaluation CAVE CAVE system/PhaseSpace motion capture Delmia
[39], [82] Workplace design and optimization Desktop None Unreal Engine/Adobe Fuse/Mixamo/3Ds Max
[42] Human-robot collaboration simulation Desktop Xsens motion capture system Industrial path solutions (IPS) IMMA
[44], [57] Assembly/maintenance simulation and evaluation HMD NDI POLARIS optical measurement system/CyberGlove/ Sensics HMD VESP/OSG
[48] Assembly/maintenance simulation and evaluation Desktop Kinect Jack
[49] Assembly/maintenance simulation and evaluation Desktop None IPS IMMA
[60] Assembly/maintenance simulation and evaluation Both Oculus Rift/Vicon optical tracking system/Tablet computer Unity 3D/Middle VR/ Vuforia
[61] Assembly/maintenance simulation and evaluation CAVE CAVE system/Kinect/Firefly cameras Jack/VR Juggler/OpenGL
[63] Automotive interior assessment AR Iphone/VITUS XXL laser scanner Apple ARKit/UMTRI human shape/Unity 3D
[64] Automotive interior assessment Desktop None IPS IMMA
[65] Automotive interior assessment AR Xsens motion capture system/ Microsoft HoloLens CATIA/Blender/Unity
[72], [75] Workplace design and optimization Both Vicon optical tracking system/ Volfoni Edge RF/AVR sound system Delmia
[80] Human behavior study Desktop None RAMSIS
[83] Human-robot collaboration simulation HMD Oculus Rift/Kinect Unity 3D/Evolver 3D digital avatar/3Ds Max
Most desktop application environments do not require additional hardware outside of ubiquitous devices and the settings and simulations for DHMs can be implemented in various software using a keyboard and mouse. However, for CAVE and HMD application environments, DHM simulations are driven by motion data captured from real humans. Additionally, some studies have utilized motion capture methods instead of a keyboard and mouse to drive DHMs in desktop environments[34,42,48,56]. Commonly used motion capture systems can be divided into two types: non-optical systems and optical systems. As shown in Figure 10, Xsens is the most common non-optical system. It captures the human motion using wearable inertial sensors attached to the human body. Optical systems include non-marker (e.g., Kinect) and marker-based (e.g., Vicon and NDI) methods. Marker-based optical systems must fix several markers on the human body and perform system calibration, which is a complex and difficult process. Non-optical and marker-based optical systems are typically expensive.
Additionally, wearing sensors or markers can easily make users feel uncomfortable and influence the realism of human behavior[72]. Currently, Kinect is the most commonly used and economical motion capture system. It was originally developed for use in games. Although the accuracy of Kinect is not as high as that of other motion capture systems, some scholars have improved its accuracy by utilizing multiple Kinect systems, which can meet the requirements of applications in manufacturing industries[48,79]. For HMD systems, different types of helmets are used, such as the Oculus Rift, Sensics, and eMagin[44,54,57,60,78,83]. For AR systems, portable devices, such as the Microsoft HoloLens, tablets, and smart phones, are commonly used hardware[60,63,65].
In terms of software, Jack, RAMSIS, and Delmia are the most commonly used commercial software for DHM applications. RAMSIS provides designers with detailed DHMs to simulate various operation behaviors of occupants. It is an efficient tool for the HFE analysis and evaluation of automotive interiors[80]. Jack and Delmia are powerful tools for DHMs and simulations that can analyze and evaluate HFE issues in various stages, including product design and manufacturing. Furthermore, in recent years, game engines (GEs) have been applied to virtual manufacturing and DHMs, including Unity 3D and Unreal Engine[35,39]. GEs can provide helpful application programming interfaces for rendering, interaction, and physics, facilitating the construction of high-quality and high-fidelity virtual environments[39,82]. The DHMs used in GEs typically come from online character animation databases, such as SmartBody, Mixamo Fuse, and UMTRI Human Shapes[82,84,85]. Currently, such DHMs are mainly used to perform certain postures or movements and do not have built-in functionality for HFE analysis and evaluation. The DHMs used in the software platforms discussed above are presented in Figure 11.
3.5 Motion control
Motion control refers to the methods by which DHMs perform desired postures and movements under certain constraints. In the application of DHMs, it is sometimes only necessary to adjust DHMs to a set of specific postures. For example, for automotive seat assessment, DHMs remain seated and immobile in various sitting postures and HFE analysis and evaluation are completed based on these static postures[63]. However, in most cases, DHMs must perform continuous movements. The authenticity of such movements is crucial for subsequent HFE analysis and evaluation[48]. Although some simplifications are implemented compared to a real human body, DHMs typically have at least 20 DOFs (degree of freedom). The generation of postures and movements for DHMs is a complex process, making motion control a key technology and research focus for the application of DHMs. Generally speaking, there are five types of motion control methods for DHMs: key frame methods, motion capture methods, model-driven methods, motion synthesis methods, and motion planning methods. A comparison of these five types of methods is provided in Table 5.
Comparison of motion control methods
Method Reality Controllability Efficiency Universality Article
Key frame Good Good Poor Good

[38][39][40][50][51][52][58][59][62][63][64][66]

[67][68][69][71][72][73][74][75][76][80][81][82]

Motion capture General General Good General

[33][34][35][37][41][42][44][48][52][54][55][56]

[57][60][61][65][70][72][75][77][78][79][83]

Model-driven Good Poor Good Poor [52][44]
Motion synthesis General General General Poor [84]
Motion planning General Poor Good Good [49][53]
In key frame methods, designers manually adjust each key posture of a DHM and interpolate key postures to generate continuous movements, as shown in Figure 12a. Designers must have a very high level of ergonomics expertise and experience to coordinate the many degrees of freedom of DHMs. This type of method is inefficient and requires a large amount of work for complex environments and tasks[48]. However, because designers can control every joint of a DHM, this type of method has good controllability and universality.
Motion capture methods drive DHMs using motion data captured from real humans and have the highest level of efficiency. However, the application scope of motion capture methods is limited. It is currently difficult to complete movements that require interaction between DHMs and environments, such as the process of tightening bolts[44]. As shown in Figure 12b, users complete various postures and movements without real objects, meaning the resulting postures and movements are inaccurate based on a lack of real interaction. Therefore, captured motion data cannot accurately reflect human motions, meaning the motion accuracy of DHMs driven in this manner is relatively low. To overcome this issue, Peruzzini et al.[72,75] established a mixed immersive environment by combining virtual and real objects, as shown in Figure 7c. A mixed environment makes user postures and movements more natural and real, improving the motion accuracy of DHMs.
Model-driven methods construct motion functions for human motions, then generate continuous movements by inputting different parameters into the motion functions. However, this method can only be used for certain regular behaviors, such as walking[52] and tightening operations[44].
Motion synthesis methods collect and store a large amount of human motion data in advance. Then, when handling a new task, stored motion data is edited and synthesized to generate the desired motions for DHMs, as shown in Figure 12c. However, human motion editing and synthesis technology is complex and problem solving efficiency is low[86]. Furthermore, because these methods mainly utilize stored motion data, their universality is very poor.
Motion planning methods are used to solve the problem of motion generation for DHMs in complex environments and tasks and are mainly applied in the field of robotics. For certain complex environments and tasks, the four previous methods often struggle to obtain reasonable postures and movements for DHMs, but motion planning is able to solve such problems. For example, in Figure 12d, the purpose of the DHM simulation is to verify whether or not workers can complete assembly operations with their arm inside a narrow gap. This problem cannot be solved using model-driven methods, but could be solved using key frame or motion synthesis methods, but the resulting workload would be very large. Motion capture methods can also solve this problem but would require excessive human and material resources. In this case, motion planning methods are the most suitable choice. However, based on the use of random sampling in motion planning methods, the results of motion planning are always random and different. Additionally, the realism of generated motions is poor and requires further optimization.
Based on statistical results, key frame and motion capture methods are the main motion control methods used in manufacturing industries. In contrast, model-driven methods, motion synthesis methods, and motion planning methods are rarely used. Additionally, some studies have combined different methods to improve the efficiency and accuracy of motion control. For example, Qiu et al.[44] combined model-driven and motion capture methods. Guo et al.[52] combined key frame methods, model-driven methods, and motion capture methods.
3.6 HFE evaluation
Based on control by human subjects (designers), DHMs can complete various postures and movements and realize interaction with virtual environments. This situation can also be viewed as human subjects interacting with virtual environments. Therefore, HFE issues can be evaluated from the perspectives of both human subjects and DHMs[72,75]. Accordingly, HFE evaluation can be divided into two types: DHM-based methods and human-subject-based methods. The former type conducts HFE evaluation by analyzing simulation processes based on DHMs, while the latter type focuses on interviews or questionnaires completed by human subjects[58].
The factors considered in DHM-based methods mainly focus on reachability, visibility, postures, comfort, fatigue, and operation space. Major analysis tools and methods include the envelope ball, vision cone, RULA[37,42,44], rapid entire body assessment[74,76,79], Ovako working posture assessment system[76,81], and European assessment worksheet[34,71], which are typically integrated into commercial software. Human-subject-based methods mainly focuses on factors that cannot be evaluated using simulations. For example, operation difficulty and stress are difficult to quantify using simulations, but they can be measured based on the subjective scores of human subjects. Commonly used methods include HFE checklists and the NASA task load index questionnaire[12,60]. Additionally, the content of interviews and questionnaires can be varied according to specific tasks.
4 Research trends and conclusions
In this section, various research trends are analyzed and discussed by answering research question Q4. Next, the work performed in this study is summarized.
4.1 Research trends
Based on our analysis of the articles discussed above, we believe that research on DHMs in manufacturing industries will progress based on the following trends.
4.1.1   Human behavior studies in manufacturing processes
Human behavior during manufacturing processes has a significant influence on production efficiency, which requires further study[33]. With the continuous improvement of GEs and computer simulation technology, it is becoming easier to construct high-quality and high-fidelity interactive virtual environments[39]. Therefore, different manufacturing scenarios can be constructed in advance to study various worker behaviors in immersive virtual environments. Some scholars have begun to study questions related to the peer effect[33,40,80], but additional factors should be considered in the future, such as the different physical characteristics of DHMs and affective communication or dialogue among DHMs. Furthermore, cognitive DHMs can be introduced into manufacturing processes to create more realistic manufacturing environments. The realistic nature of virtual environments makes acquired data and information more credible. Additionally, simulation results are not limited to the research level, but can also be used as a basis to improve working scenarios and content directly.
4.1.2   Planning and assessment of human-robot collaboration systems
Human-robot collaboration is a major trend in manufacturing industries. In the future, a large number of manual stations will be transformed into human-robot collaboration stations[42]. Additionally, based on the high flexibility of human-robot collaboration, human-robot collaboration stations should be adjusted frequently to meet different production needs. The rational design of human-robot collaboration systems is very important, and virtual simulation and DHMs are considered to be key technologies for this design process[35]. In traditional applications, the interactive objects for DHMs are mostly static environments and products. However DHMs must interact with dynamic robots in state-of-the-art applications. Therefore, future research should first focus on reflecting the features of human-robot interactions in the physical world in a virtual simulation environment. For example, virtual collaborative robots should also be able to detect the positions and postures of DHMs in real time, then perform corresponding actions. Furthermore, during planning, the integration of task allocation, operation sequences, security, and efficiency should be studied. In some cases, the mental workload of workers in a dynamic interactive environment should also be considered. Otherwise, subsequent interactions may be adversely affected. There are currently no practical and effective assessment systems for human-robot collaboration systems, so future studies should consider how to evaluate human-robot collaboration systems effectively and practically.
4.1.3   Combination with AR
In the absence of 3D models or in cases where models are complex and the modeling process is very difficult, some HFE issues can be analyzed by combining AR and DHMs[63,65]. However, current DHMs are very static and can only have limited interactions with physical environments. In the future, intelligent interactions between DHMs and physical environments should be studied. This will allow DHMs to generate postures and movements automatically according to different needs while supporting HFE analysis and evaluation. Additionally, this combination can be applied to education and training in manufacturing industries.
4.1.4 Improved motion planning for DHMs
Based on the popularity of DHMs, an increasing number of researchers will consider DHMs in their studies. Motion control will always be a key technology for the application of DHMs in manufacturing industries. Currently, the most commonly used methods are key frame methods and motion capture methods. However, based on growth in labor costs and the trends of smart manufacturing, these two methods, which rely heavily on manual intervention, will no longer be adequate[49]. Particularly in complex constrained environments, motion planning will be the most appropriate method. Current solutions for motion planning are similar to robotic motions, which are based on random sampling methods. However, unlike robots, human motions have unique rules and characteristics. Therefore, the laws of human motion in different environments should be studied in the future to guide the sampling process and improve planning efficiency.
4.1.5   Systematical and comprehensive HFE evaluation
HFE evaluation is very important and serves as the basis of product design and improvement. Existing HFE evaluation systems are often simplified by ignoring various influencing factors or considering only common manufacturing conditions[72]. However, as product structures and working environments become more complex, HFE issues should be systematically and comprehensively analyzed and evaluated. First, various manufacturing conditions should be considered, including different lighting environments[57], heights and ages of workers[80], and duration of work[62]. Based on HFE evaluations of different conditions, the most appropriate designs and improvement schemes can be identified. Second, additional factors or indexes should be considered. For example, for DHM-based methods, more detailed DHMs or tools, such as AnyBody[87], should be utilized to derive more useful information from simulations.
4.2 Conclusions
Based on the development of VR, virtual manufacturing and DHMs have been widely applied and developed in recent years. This paper aimed to identify the applications and research trends of DHMs in manufacturing industries. First, related articles from the past five years (January of 2014 to May of 2019) were searched and screened. A final set of 49 articles was obtained. Next, based on these articles, the applications of DHMs in manufacturing industries were analyzed based on six aspects: application field, application type, application environment, development tools, motion control, and HFE evaluation. Finally, certain research trends were illustrated and discussed based on various methods of analysis. The results indicate that applications of DHMs in manufacturing should consider the following aspects: (1) Human behavior studies in manufacturing processes; (2) Planning and assessment of human-robot collaboration systems; (3) Combinations with AR; (4) Improved motion planning for DHMs; and (5) Systematical and comprehensive HFE evaluation.
To summarize, the application of DHMs has excellent prospects in manufacturing industries and this paper should provide a helpful reference for readers engaged in DHM-related research.

Reference

1.

Ma J F, Jaradat R, Ashour O, Hamilton M, Jones P, Dayarathna V L. Efficacy investigation of virtual reality teaching module in manufacturing system design course. Journal of Mechanical Design, 2019, 141(1): 1–13 DOI:10.1115/1.4041428

2.

Nee A Y C, Ong S K. Virtual and augmented reality applications in manufacturing. IFAC Proceedings Volumes, 2013, 46(9): 15–26 DOI:10.3182/20130619-3-ru-3018.00637

3.

Jensen L, Konradsen F. A review of the use of virtual reality head-mounted displays in education and training. Education and Information Technologies, 2018, 23(4): 1515–1529 DOI:10.1007/s10639-017-9676-0

4.

Saposnik G, Levin M. Virtual reality in stroke rehabilitation. Stroke, 2011, 42(5): 1380–1386 DOI:10.1161/strokeaha.110.605451

5.

Haydar M, Roussel D, Maïdi M, Otmane S, Mallem M. Virtual and augmented reality for cultural computing and heritage: a case study of virtual exploration of underwater archaeological sites (preprint). Virtual Reality, 2011, 15(4): 311–327 DOI:10.1007/s10055-010-0176-4

6.

Azizi A, Ghafoorpoor Yazdi P, Hashemipour M. Interactive design of storage unit utilizing virtual reality and ergonomic framework for production optimization in manufacturing industry. International Journal on Interactive Design and Manufacturing (IJIDeM), 2019, 13(1): 373–381 DOI:10.1007/s12008-018-0501-9

7.

Pohit G, Kumar K. Virtual manufacturing of various types of gears and validation of the technique using rapid prototype. Virtual and Physical Prototyping, 2012, 7(2): 153–171 DOI:10.1080/17452759.2012.686696

8.

Karvouniari A, Michalos G, Dimitropoulos N, Makris S. An approach for exoskeleton integration in manufacturing lines using Virtual Reality techniques. Procedia CIRP, 2018, 78: 103–108 DOI:10.1016/j.procir.2018.08.315

9.

Bougaa M, Bornhofen S, Kadima H, Rivière A. Virtual reality for manufacturing engineering in the factories of the future. Applied Mechanics and Materials, 2015, 789/790: 1275–1282 DOI:10.4028/www.scientific.net/amm.789-790.1275

10.

Guo Z Y, Zhou D, Chen J Y, Geng J, Lv C, Zeng S K. Using virtual reality to support the product's maintainability design: Immersive maintainability verification and evaluation system. Computers in Industry, 2018, 101: 41–50 DOI:10.1016/j.compind.2018.06.007

11.

Vergnano A, Berselli G, Pellicciari M. Interactive simulation-based-training tools for manufacturing systems operators: an industrial case study. International Journal on Interactive Design and Manufacturing (IJIDeM), 2017, 11(4): 785–797 DOI:10.1007/s12008-016-0367-7

12.

Aromaa S, Frangakis N, Tedone D, Viitaniemi J, Aaltonen I. Digital human models in human factors and ergonomics evaluation of gesture interfaces. Proceedings of the ACM on Human-Computer Interaction, 2018, 2(EICS): 1–14 DOI:10.1145/3229088

13.

So R H Y, Lam S T. Factors affecting the appreciation generated through applying human factors/ergonomics (HFE) principles to systems of work. Applied Ergonomics, 2014, 45(1): 99–109 DOI:10.1016/j.apergo.2013.04.019

14.

Guo Z Y, Zhou D, Liu P Y, He Z Y, Lv C. A quantitative assessment method for the space design of products based on ergonomics and virtual simulation. PLoS One, 2018, 13(7): e0200880 DOI:10.1371/journal.pone.0200880

15.

Chen M, Liu J F. Virtual simulation of production line for ergonomics evaluation. Advances in Manufacturing, 2014, 2(1): 48–53 DOI:10.1007/s40436-014-0060-7

16.

Rossoni M, Bergonzi L, Colombo G. Integration of virtual reality in a knowledge-based engineering system for preliminary configuration and quotation of assembly lines. Computer-Aided Design and Applications, 2018, 16(2): 329–344 DOI:10.14733/cadaps.2019.329-344

17.

Gao W, Shao X D, Liu H L. Virtual assembly planning and assembly-oriented quantitative evaluation of product assemblability. The International Journal of Advanced Manufacturing Technology, 2014, 71(1/2/3/4): 483–496 DOI:10.1007/s00170-013-5514-8

18.

Plantard P, Shum H P H, Le Pierres A S, Multon F. Validation of an ergonomic assessment method using Kinect data in real workplace conditions. Applied Ergonomics, 2017, 65: 562–569 DOI:10.1016/j.apergo.2016.10.015

19.

Qiu S G, Fan X M, Wu D L, He Q C, Zhou D J. Virtual human modeling for interactive assembly and disassembly operation in virtual reality environment. The International Journal of Advanced Manufacturing Technology, 2013, 69(9/10/11/12): 2355–2372 DOI:10.1007/s00170-013-5207-3

20.

Chang S W, Wang M J J. Digital human modeling and workplace evaluation: Using an automobile assembly task as an example. Human Factors and Ergonomics in Manufacturing, 2007, 17(5): 445–455 DOI:10.1002/hfm.20085

21.

Chaffin D B. Human motion simulation for vehicle and workplace design. Human Factors and Ergonomics in Manufacturing, 2007, 17(5): 475–484 DOI:10.1002/hfm.20087

22.

Makarov S N, Noetscher G M, Yanamadala J, Piazza M W, Louie S, Prokop A, Nazarian A, Nummenmaa A. Virtual human models for electromagnetic studies and their applications. IEEE Reviews in Biomedical Engineering, 2017, 10: 95–121 DOI:10.1109/rbme.2017.2722420

23.

Chaffin D B. Digital human modeling for workspace design. Reviews of Human Factors and Ergonomics, 2008, 4(1): 41–74 DOI:10.1518/155723408x342844

24.

Pearson F. Systematic approaches to a successful literature review. Educational Psychology in Practice, 2014, 30(2): 205–206 DOI:10.1080/02667363.2014.900913

25.

dos Santos A C C, Delamaro M E, Nunes F L S. The relationship between requirements engineering and virtual reality systems: a systematic literature review. In: Proceedings of the 15th Symposium on Virtual and Augmented Reality, Cuiaba, Mato Grosso, Brazil, IEEE, 2013, 53–62 DOI: 10.1109/SVR.2013.52

26.

Berg L P, Vance J M. Industry use of virtual reality in product design and manufacturing: a survey. Virtual Reality, 2017, 21(1): 1–17 DOI:10.1007/s10055-016-0293-9

27.

Neumann D L, Moffitt R L, Thomas P R, Loveday K, Watling D P, Lombard C L, Antonova S, Tremeer M A. A systematic review of the application of interactive virtual reality to sport. Virtual Reality, 2018, 22(3): 183–198 DOI:10.1007/s10055-017-0320-5

28.

Li X, Yi W, Chi H L, Wang X Y, Chan A P C. A critical review of virtual and augmented reality (VR/AR) applications in construction safety. Automation in Construction, 2018, 86: 150–162 DOI:10.1016/j.autcon.2017.11.003

29.

Palmarini R, Erkoyuncu J A, Roy R, Torabmostaedi H. A systematic review of augmented reality applications in maintenance. Robotics and Computer-Integrated Manufacturing, 2018, 49: 215–228 DOI:10.1016/j.rcim.2017.06.002

30.

Mochimaru M,. Digital human models for human-centered design. Journal of Robotics and Mechatronics, 2017, 29(5): 783–789 DOI:10.20965/jrm.2017.p0783

31.

Eiris R, Gheisari M. Research trends of virtual human applications in architecture, engineering and construction. Electronic Journal of Information Technology in Construction, 2017, 22(22): 168–184

32.

Sanjog J, Karmakar S, Patel T, Chowdhury A. Towards virtual ergonomics: aviation and aerospace. Aircraft Engineering and Aerospace Technology, 2015, 87(3): 266–273 DOI:10.1108/aeat-05-2013-0094

33.

GürerkÖ, Bönsch A, Kittsteiner T, Staffeldt A. Virtual humans as co-workers: A novel methodology to study peer effects. Journal of Behavioral and Experimental Economics, 2019, 78: 17–29 DOI:10.1016/j.socec.2018.11.003

34.

Nikolakis N, Alexopoulos K, Xanthakis E, Chryssolouris G. The digital twin implementation for linking the virtual representation of human-based production tasks to their physical counterpart in the factory-floor. International Journal of Computer Integrated Manufacturing, 2019, 32(1): 1–12 DOI:10.1080/0951192x.2018.1529430

35.

Matsas E, Vosniakos G C. Design of a virtual reality training system for human–robot collaboration in manufacturing tasks. International Journal on Interactive Design and Manufacturing (IJIDeM), 2017, 11(2): 139–153 DOI:10.1007/s12008-015-0259-2

36.

Bures M. Time demands of virtual ergonomic modelling-experimental study. In: Proceedings of the 26th International Symposium on Intelligent Manufacturing and Automation, Croatia, 2015, 286–291 DOI:10.2507/26th.daaam.proceedings.039

37.

Wang W, Feng W J. The research of maintainability analysis based on immersive virtual maintenance technology. In: Proceedings of the 2017 International Conference on Human Factors in Simulation and Modeling, Los Angeles, United States, AHFE, 2017, 573–582 DOI:10.1007/978-3-319-60591-3_52

38.

Geng J, Peng X, Li Y, Lv C, Wang Z L, Zhou D. A semi-automatic approach to implement rapid non-immersive virtual maintenance simulation. Assembly Automation, 2018, 38(3): 291–302 DOI:10.1108/aa-07-2017-079

39.

Paravizo E, Braatz D. Using a game engine for simulation in ergonomics analysis, design and education: An exploratory study. Applied Ergonomics, 2019, 77: 22–28 DOI:10.1016/j.apergo.2019.01.001

40.

Li M K, Liu Y K, Peng M J, Xie C L, Yang L Q. The digital simulation and fuzzy evaluation to reduce the likelihood of unsafe behavior in nuclear decommissioning. Annals of Nuclear Energy, 2018, 119: 331–341 DOI:10.1016/j.anucene.2018.05.023

41.

Wang Y Y, Zhou B, Shao H W, Peng X. Study on the method of intelligent maintenance simulation based on the fusion of human motion data. In: Proceedings of the 5th International Conference on Computer Systems, Electronics and Control, Dalian, China, IEEE, 2017, 323–328 DOI:10.1109/ICCSEC.2017.8446910

42.

Castro P R, Högberg D, Ramsen H, Bjursten J, Hanson L. Virtual simulation of human-robot collaboration workstations. In: Proceedings of the 20th Congress of the International Ergonomics Association, Florence, Italy, Springer, 2019, 250–261DOI:10.1007/978-3-319-96077-7_26

43.

Geng J, Li Y, Wang R R, Wang Z L, Lv C, Zhou D. A virtual maintenance-based approach for satellite assembling and troubleshooting assessment. Acta Astronautica, 2017, 138: 434–453 DOI:10.1016/j.actaastro.2017.06.018

44.

Qiu S G, He Q C, Fan X M, Wu D L. Virtual human hybrid control in virtual assembly and maintenance simulation. International Journal of Production Research, 2014, 52(3): 867–887 DOI:10.1080/00207543.2013.842025

45.

Ghandi S, Masehian E. Review and taxonomies of assembly and disassembly path planning problems and approaches. Computer-Aided Design, 2015, 67: 58–86 DOI:10.1016/j.cad.2015.05.001

46.

Chung C, Peng Q J. Tool selection-embedded optimal assembly planning in a dynamic manufacturing environment. Computer-Aided Design, 2009, 41(7): 501–512 DOI:10.1016/j.cad.2009.03.007

47.

Chung C, Peng Q J. A novel approach to the geometric feasibility analysis for fast assembly tool reasoning. The International Journal of Advanced Manufacturing Technology, 2006, 31(1/2): 125–134 DOI:10.1007/s00170-005-0173-z

48.

Jun C M, Lee J Y, Kim B H, Noh S D. Automatized modeling of a human engineering simulation using Kinect. Robotics and Computer-Integrated Manufacturing, 2019, 55: 259–264 DOI:10.1016/j.rcim.2018.03.014

49.

Li Y, Delfs N, Mårdberg P, Bohlin R, Carlson J S. On motion planning for narrow-clearance assemblies using virtual manikins. Procedia CIRP, 2018, 72: 790–795 DOI:10.1016/j.procir.2018.03.181

50.

Y F Y, Xue Q, Liu M X. Virtual human motion design and ergonomics analysis in maintenance simulation. In: Proceedings of the International Conference on Digital Human Modeling and Simulation. Walt Disney World, United States, 2017, 65–74 DOI:10.1007/978-3-319-41627-4_7

51.

Song Q, Zhang J D, Li C P, Wang Z Q. Application of virtual simulation technology in maintenance training. In: Proceedings of the 11th International Conference on E-Learning and Games, Edutainment. Bournemouth, United Kingdom, 2017, 101–107 DOI:10.1007/978-3-319-65849-0_12

52.

Guo Z Q, Lv C, Zhou D, Peng X, Wang Z L. Mixing control of animating virtual human for maintenance simulation. In: Proceedings of the 12th World Congress on Intelligent Control and Automation, Guilin, China, IEEE, 2016, 1091–1098 DOI:10.1109/WCICA.2016.7578807

53.

Mårdberg P, Yan Y, Bohlin R, Delfs N, Gustafsson S, Carlson J S. Controller hierarchies for efficient virtual ergonomic assessments of manual assembly sequences. In: Proceedings of the 6th CIRP Conference on Assembly Technologies and Systems. Gothenburg, Sweden, 2016, 435–440 DOI:10.1016/j.procir.2016.02.084

54.

Bönig J, Fischer C, Weckend H, Döbereiner F, Franke J. Accuracy and immersion improvement of hybrid motion capture based real time virtual validation. In: Proceedings of the 24th CIRP Design Conference. Pirelli Lombardy Region, Italy, 2014, 294–299 DOI: 10.1016/j.procir.2014.03.191

55.

Louison C, Ferlay F, Keller D, Mestre, D R. Operators’ accessibility studies for assembly and maintenance scenarios using virtual reality. Fusion Engineering and Design, 2017, 124: 610–614 DOI:10.1016/j.fusengdes.2017.03.017

56.

Xu Y J, Hou W K. Calculation of operational domain of virtual maintenance based on convex hull algorithm. In: Proceedings of the 2th International Conference on Reliability Systems Engineering, Beijing, China, 2017DOI: 10.1109/ICRSE.2017.8030789

57.

He Q C, Qiu S G, Fan X M, Liu K Y. An interactive virtual lighting maintenance environment for human factors evaluation. Assembly Automation, 2016, 36(1): 1–11 DOI:10.1108/AA-04-2015-029

58.

Guo J W, Sun Z Z, He J X, Jia X J, Li H J, Yan X H, Chen H B, Tang H, Wu G H. An approach for integrated analysis of human factors in remote handling maintenance. Science and Technology of Nuclear Installations, 2016, 1–11 DOI:10.1155/2016/9108751

59.

Alkan B, Vera D, Ahmad M, Ahmad B, Harrison R. A lightweight approach for human factor assessment in virtual assembly designs: an evaluation model for postural risk and metabolic workload. Procedia CIRP, 2016, 44: 26–31 DOI:10.1016/j.procir.2016.02.115

60.

Aromaa S, Väänänen K. Suitability of virtual prototypes to support human factors/ergonomics evaluation during the design. Applied Ergonomics, 2016, 56: 11–18 DOI:10.1016/j.apergo.2016.02.015

61.

Puthenveetil S C, Daphalapurkar C P, Zhu W J, Leu M C, Liu X F, Gilpin-Mcminn J K, Snodgrass S D. Computer-automated ergonomic analysis based on motion capture and assembly simulation. Virtual Reality, 2015, 19(2): 119–128 DOI:10.1007/s10055-015-0261-9

62.

Geng J, Lv C, Zhou D, Li Y, Wang Z L. Compensation-based methodology for maintenance time prediction in a virtual environment. Simulation Modelling Practice and Theory, 2014, 47: 92–109 DOI:10.1016/j.simpat.2014.05.008

63.

Park B K Daniel, Reed M P. Accommodation assessments for vehicle occupants using augmented reality. In: Proceedings of the 20th Congress of the International Ergonomics Association. Florence, Italy, Springer, 2019, 3–9 DOI:10.1007/978-3-319-96077-7_1

64.

Högberg D, Castro P R, Mårdberg P, Delfs N, Nurbo P, Fragoso P, Andersson L, Brolin E, Hanson L. DHM based test procedure concept for proactive ergonomics assessments in the vehicle interior design process. Advances in Intelligent Systems and Computing. Florence, Italy, Springer, 2018: 314–323 DOI:10.1007/978-3-319-96077-7_33

65.

Wan J, Wang N X. A method of motion-based immersive design system for vehicle occupant package. In: Proceedings of ASME Conference on ASME 2018 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. Quebec, Canada, ASME, 2018, 1–6 DOI:10.1115/DETC2018-85054

66.

Deng L, Wang G H, Chen B. Operating comfort prediction model of human-machine interface layout for cabin based on GEP. Computational Intelligence and Neuroscience, 2015, 1–13 DOI:10.1155/2015/896072

67.

Stefania S, Danila G, Fabrizio S, Ghibaudo L. FCA ergonomics proactive approach in developing new cars: virtual simulations and physical validation. Advances in Intelligent Systems and Computing. Florence, Italy, Springer, 2016, 57–63 DOI:10.1007/978-3-319-41627-4_6

68.

Niu H Y, Li M H, Yang Y. Modeling and application virtual human for ergonomics evaluation of armored vehicle. In: Proceedings of the 13th International Conference on Man-Machine-Environment System Engineering. Yantai, China, Springer, 2014, 111–120 DOI:10.1007/978-3-642-38968-9_13

69.

Mircheski I, Kandikjan T, Sidorenko S. Comfort analysis of vehicle driver's seat through simulation of the sitting process. Tehnicki Vjesnik, 2017, 21(2): 291–298 DOI:hrcak.srce.hr/120380

70.

Tao Q, Kang J S, Sun W L, Li Z B, Huo X. Digital evaluation of sitting posture comfort in human-vehicle system under industry 4.0 framework. Chinese Journal of Mechanical Engineering, 2016, 29(6): 1096–1103 DOI:10.3901/cjme.2016.0718.082

71.

Caputo F, Greco A, Fera M, Caiazzo G, Spada S. Simulation techniques for ergonomic performance evaluation of manual workplaces during preliminary design phase. Advances in Intelligent Systems and Computing. Florence, Italy, Springer, 2019, 170–180 DOI:10.1007/978-3-319-96077-7_18

72.

Peruzzini M, Pellicciari M, Gadaleta M. A comparative study on computer-integrated set-ups to design human-centred manufacturing systems. Robotics and Computer-Integrated Manufacturing, 2019, 55: 265–278 DOI:10.1016/j.rcim.2018.03.009

73.

Mårdberg P, Fredby J, Engstrom K, Li Y, Bohlin R. A novel tool for optimization and verification of layout and human logistics in digital factories. In: Proceedings of the 51st CIRP Conference on Manufacturing Systems. Sweden, 2018, 545–550 DOI:10.1016/j.procir.2018.03.158

74.

Karmakar S, Solomon R. Ergonomic evaluations and design interventions for shop-floors dealing with chemical conversion coatings: case study from India. Advances in Ergonomics in Design. Los Angeles, United States, AHFE, 2017, 857–868 DOI:10.1007/978-3-319-60582-1_87

75.

Peruzzini M, Carassai S, Pellicciari M. The benefits of human-centred design in industrial practices redesign of workstations in pipe industry. In: Proceedings of the 27th International Conference on Flexible Automation and Intelligent Manufacturing, Modena, Italy, 2017, 11, 1247–1254DOI:10.1016/j.promfg.2017.07.251

76.

Sanjog J, Patnaik B, Patel T, Karmakar S. Context-specific design interventions in blending workstation: an ergonomics perspective. Journal of Industrial and Production Engineering, 2016, 33(1): 32–50 DOI:10.1080/21681015.2015.1099057

77.

Maurice P, Padois V, Measson Y, Bidaud P. Human-oriented design of collaborative robots. International Journal of Industrial Ergonomics, 2017, 57: 88–102 DOI:10.1016/j.ergon.2016.11.011

78.

Matsas E, Vosniakos G C, Batras D. Modelling simple human-robot collaborative manufacturing tasks in interactive virtual environments. In: Proceedings of the 2016 Virtual Reality International Conference. Laval, France, ACM Press, 2016DOI: 10.1145/2927929.2927948

79.

Thomas C, Stankiewicz L, Grötsch A, Wischniewski S, Deuse J, Kuhlenkötter B. Intuitive work assistance by reciprocal human-robot interaction in the subject area of direct human-robot collaboration. Procedia CIRP, 2016, 44: 275–280 DOI:10.1016/j.procir.2016.02.098

80.

Wirsching H J, Spitzhirn M. Virtual aging–Implementation of age-related human performance factors in ergonomic vehicle design using the digital human model RAMSIS. Advances in Intelligent Systems and Computing. Florence, Italy, Springer, 2019, 87–97 DOI:10.1007/978-3-319-96065-4_12

81.

Lanzotti A, Tarallo A, Carbone F, Coccorese D, D’Angelo R, di Gironimo G, Grasso C, Minopoli V, Papa S. Interactive tools for safety 4.0: virtual ergonomics and serious games in tower automotive. Advances in Intelligent Systems and Computing. Florence, Italy, Springer, 2019, 270–280 DOI:10.1007/978-3-319-96077-7_28

82.

Paravizo E, Braatz D. Employing game engines for ergonomics analysis, design and education. Advances in Intelligent Systems and Computing. Florence, Italy, Springer, 2019, 330–338 DOI:10.1007/978-3-319-96077-7_35

83.

Zhang D G, Li Z H, Xiao Y W, Jiang L, Wang D D, Wang G P. Design and Implementation of Simulation System for Safety Accident Case Based on Immersive Virtual Reality Technology. In: Proceedings of the 3rd Annual International Conference on Electronics, Electrical Engineering and Information Science. Guangzhou, China, Atlantis Press, 2017, 131: 416–420 DOI:10.2991/eeeis-17.2017.58

84.

SmartBody. http://smartbody.ict.usc.edu/

85.

UMTRI Human Shapes. http://humanshape.org/

86.

Gaisbauer F, Lehwald J, Agethen P, Otto M, Rukzio E. A motion reuse framework for accelerated simulation of manual assembly processes. Procedia CIRP, 2018, 72: 398–403 DOI:10.1016/j.procir.2018.03.282

87.

Chander D S, Cavatorta M P. Multi-directional one-handed strength assessments using AnyBody Modeling Systems. Applied Ergonomics, 2018, 67: 225–236 DOI:10.1016/j.apergo.2017.09.015