Design and Development of a Virtual Reality Training Platform for Fiber-Reinforced Composite Manufacturing

doi:10.21203/rs.3.rs-4118899/v1

Download PDF

Research Article

Design and Development of a Virtual Reality Training Platform for Fiber-Reinforced Composite Manufacturing

https://doi.org/10.21203/rs.3.rs-4118899/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

This study presents the design and development of a virtual reality (VR) training platform for manufacturing fiber-reinforced composites, a sophisticated and high-demand material in various industries. Due to the high costs and safety concerns associated with compression molding machines, essential equipment for this manufacturing process, the VR platform offers a promising alternative for training in educational institutions and manufacturing industries. The platform provides immersive and interactive training in a cost-efficient learning environment, allowing users to gain practical experience without the risks and expenses of physical training. The VR training module was developed using the Unity game engine and deployed on the HTC Vive Pro headset. While virtual reality-based training has become very popular in both educational and industrial settings, very few studies have confirmed the knowledge transfer from immersive training to real-world tasks. In this experimental study, researchers compared the performance of participants who received prior VR training with those who did not for their real-world performance in manufacturing. The findings of this research have the potential to significantly impact the effectiveness and transferability of virtual-reality-based training in composite manufacturing. Educational institutions and manufacturing industries can utilize the VR training platform to provide a safer and more cost-effective training environment for manufacturing fiber-reinforced composites.

Virtual Reality

Fiber-reinforced Composites

Digital Learning

VR Training

Fiber-reinforced composites have recently become one of the market's most sophisticated and demanding materials (Rajak et al. 2019). The expected market growth of composite material is 6.1% from 2021 to 2028 and is predicted to reach 136.5 billion USD by 2028 (Data Bridge Market Research, 2021). The lightweight and high-strength properties of these materials have brought much popularity to them in diverse applications, thus increasing the need for expertise in the manufacturing process. Compression molding is an efficient and leading-edge manufacturing process of fiber-reinforced composite (Stelzer et al., 2022). Though a highly competitive manufacturing process, compression molding machines are usually expensive because of their high holding cost and energy expenditure (Rawi, Alias, and Wahid 2020). Compression molding dominates the composite industry but is not readily accessible for reaching the learning population yet. It requires high-priced raw materials that must be stored in a controlled environment and extensive safety precautions due to the machine’s capability of high tonnage compression. The impediment of the accessibility of the process limits colleges, universities, and even manufacturing companies from practically training their students or new hires in this manufacturing process. Therefore, alternative training programs that do not involve physical resources are viable options for academic institutions and manufacturing industries.

The use of Virtual Reality (VR) as a medium for training has continuously expanded in recent years with its applications in education and training. According to the recent statistics from Business Research Company, 97% of students intend to take a VR course (Ip 2022). The global VR market growth in education is projected to reach around 13 million USD by 2026, exhibiting a compound annual growth rate of 42.9% (Fortune Business Insights 2019). VR training has been utilized to learn machine interaction and manufacturing processes by applying its extended functionalities to simulate real-world machine interactions (Jimeno and Puerta 2007). Some key advantages of using VR as a training platform include time savings, safety in learning, and meager training cost(Coburn, Freeman, and Salmon 2017) – making it a promising learning platform for high-cost manufacturing processes like compression molding of fiber-reinforced composites. While this emerging technology can offer many opportunities with reduced cost and effort, the knowledge transfer from VR training is still lacking statistical evidence (Souza et al. 2020).

The current research designed and developed a virtual compression molding machine for virtual reality-based training in the Reinforced Fiber Composites manufacturing process. The virtual platform of the study created a fast-simulated immersive training workspace of the compression molding machine. VR training provides unlimited attempts for learning as it is cost-efficient and easy to reset. This immersive and interactive training creates long-lasting learning effects for students, as VR can offer a practical learning experience (Bell and Fogler 1995). This training module was designed with a 3D game engine called Unity, a popular designing tool for virtual workspaces (Gabajová et al. 2021). The study used a head-mounted device called the HTC Vive Pro, a proven apparatus in virtual reality research (Niehorster, Li, and Lappe 2017). The Unity game engine and HTC Vive Pro combination has been numerously seen in VR training modules in the last decade (Chénéchal and Chatel-Goldman 2018; Fatima, Bukhari, and Iqbal 2022; Fang et al. 2021; Luckett 2018; Liu and Wu 2019). The VR module was designed to keep the user interaction accessible and easy to operate for its users. The module included key work components of the compression molding machine in a compact area for the users to navigate and complete the instructed tasks faster and with ease. Both quantitative and qualitative data were collected from an experimental study. After having virtual training, one group of research participants was sent to a compression molding lab to operate the physical machine. Another group of participants used the physical compression molding machine without prior virtual training. The researchers compared two groups’ performance in compression molding to investigate the knowledge transfer from virtual reality-based training.

The research had two primary goals: (1) to assess the impact of VR training on operator performance in the context of fiber-reinforced composites manufacturing, comparing it to conventional training methods and (2) to explore how the integration of VR training alongside traditional methods influences the workload experienced by individuals when performing real-world tasks.

2.1 Fiber-Reinforced Composites in Industrial Applications

Fiber-reinforced composites are engineered materials with high strength by weight and modulus by weight ratios compared to many metallic materials (Njuguna, Pielichowski, and Fan 2012). They are popular for their widespread industrial applications, especially in automotive, aviation, shipping, and other related sectors (Prashanth et al. 2017; Erden and Ho 2017). There are different types of fiber reinforcements, including glass fibers, carbon fibers, and Kevlar fibers, among many others. Carbon fibers are composed of a minimum of 92% by weight (mass fraction) carbon and can exist in the form of fibers, filaments, tows, yams, or rovings (Marsh and Rodríguez-Reinoso 2006). Carbon fibers also offer the highest specific modulus and strength in addition to their ability to retain tensile strength at high temperatures, independent of moisture (Prashanth et al. 2017). The market size for carbon fiber composite in 2021 was USD 18.4 billion and is expected to grow at a compound annual growth rate of 6% from 2022 to 2030 (Global Market Insights 2022). Because of its high industrial usage and economic importance, it is necessary to build a qualified and well-trained workforce to support production needs in the future. This research aims to solve this problem by adding a new training method using virtual reality, which has not been seen before for fiber-reinforced composite manufacturing.

2.2 Manufacturing Technique for Carbon Fiber Composite

Fiber-reinforced composites are manufactured by infusing the fibers in uncured epoxy resin and curing them (Sloan 2016). Curing is the chemical reaction between the epoxy and the hardener, resulting in a crosslinked polymer structure (Abliz et al. 2013). Prepregs are composite materials in which a high-strength reinforcement fiber is pre-impregnated with a thermoset or a thermoplastic resin (Bhatnagar and Asija 2016). Compression molding using prepregs is one of the most suitable manufacturing techniques for this type of fiber-reinforced composite (Yallew et al. 2020; Tapan Bhatt, Gohil, and Chaudhary 2018). Prepreg plies are stacked on top of each other to form a laminate, where the orientation and the number of plies depend on the desired strength and application of the material. The laminate stack is sandwiched between two mold plates and releases films to separate the epoxy from bonding to the metal mold. A release fabric known as peel ply creates a rough texture on the laminate that is often used for secondary bonding. The entire laminate layup is then placed inside a compression molding oven and cured by a recipe. The recipe can differ from machine to machine as recommended by the manufacturer. In common, there should be three segments in a typical recipe (TORAY 2020) – ramp up (increasing temperature and pressure), hold/ dwell, and ramp down (decreasing temperature and pressure). Once the recipe ends, the cured laminate is separated from the mold plates. The released films are removed and discarded, and the peel plies are peeled off. This research simulates this manufacturing technique in virtual reality training to enhance operators’ learning to perform this process successfully.

2.3 A Brief History of VR Training in Manufacturing

The use of VR in manufacturing-based industries is becoming increasingly popular (Renner et al. 2016; Doolani et al. 2020; Matsas and Vosniakos 2017; Jimeno and Puerta 2007; Matsas, Vosniakos, and Batras 2018; Choi, Jung, and Noh 2015; Salah et al. 2019; Monetti et al. 2022; Roldán et al. 2019). The design of VR training to simulate cooperation between industrial robotics manipulators and humans (Matsas and Vosniakos 2017) showed an excellent prospect of VR in human-robot applications. Similar research was conducted to observe the impact of VR training on operators of industrial robotics tasks (Monetti et al., 2022). The researchers found that participants using the VR first were quicker to complete the tasks than the others, at a higher passing rate. This training method is particularly viable because it is safe, engaging, and effective for vocational training. Doolani et al. (2020) emphasize that VR has been tested as the best immersive medium among augmented reality (AR), virtual reality, and mixed reality. Researchers have concluded in numerous examples that VR is also useful for training specific machinery. For example, in research conducted to learn the effectiveness of VR in the application for additive manufacturing process training, it was concluded that identifying and fixing a mistake in the VR is time and cost-efficient in achieving the desired quality (Renner et al. 2016). Complex hand movements to organize items in VR are also possible; for example, assembly operations. In a study conducted by Matsas and Vosniakos (2017) in prototyping proactive and adaptive techniques for human-robot collaboration, the user was able to collaborate on the lay-up of carbon fabric. Another research conducted by Matsas et al. (2018) showed that participants being trained to assemble a water pump suggested that VR, a 3D system, has a more intuitive learning approach than 2D systems. The user evaluations in VR training systems brought fruitful results in complex assemblies (Roldán et al., 2019). Users evaluated them better than a paper guide regarding mental demand, perception, and learning, improving their overall performance. VR is also a very inexpensive alternative to high-cost training. Choi et al. (2015) note that VR was previously used only in developing premier products for its low return on investment due to its high costs. However, VR has become more common in the industry because of its increased cost competitiveness. VR, having its practical applications in learning manufacturing, has inspired this research to evaluate its extended use in developing a workforce for fiber-reinforced composite manufacturing.

2.4 Effective VR System Design

Designing an effective VR system involves several key considerations to ensure that the system is user-friendly, immersive, and provides a positive user experience. The choice of hardware and the development platform is essential to ensure these features in designing and developing a virtual reality system. HTC Vive hardware system is widely used for its proven performance in precision and tracking quality (Ikbal, Ramadoss, and Zoppi 2021). Compared to similar head-mounted displays from Oculus (Currently Meta), Valve, and Samsung, the HTC Vive Pro showed balanced qualities in the main display and tracking (Angelov et al. 2020). This hardware system has been used numerous times in developing successful VR programs (Quevedo et al. 2017; Matsas and Vosniakos 2017). Unity is the most favorable development platform because of its previous application in scientific research (Dai 2022; Gabajová et al. 2021; Deb et al. 2017; Rahman et al. 2022) and provides free access for non-commercial use. Dai (2022) proposed a system to develop a simulation system for dancing using Unity. Deb et al. (2017), in their study of determining the efficacy of VR in pedestrian safety, also used Unity as the development platform. Gabajová et al. (2021) also acknowledged the effectiveness of Unity in their design and development of a 3D virtual workspace.

Different interaction methods are available in a virtual reality system, e.g., teleportation, grabbing and realizing, recasting, etc. The form of teleportation to move a player from one location to another inside the virtual environment is very effective and has a low possibility of inducing simulation sickness (Moghadam, Banigan, and Ragan 2020; Prithul, Adhanom, and Folmer 2021; Bozgeyikli et al. 2016). Using controllers to grab and release objects induces a natural feeling among users (Lin & Schulze, 2016). Ray-casting is a widespread pointing and selecting object technique that can be used in in-game menu design (Lee et al. 2003). These are the common types of interaction methods also used in this research to simulate the best user experience in attending the training.

2.5 Evaluation of Virtual Systems

Evaluation of virtual learning platforms is crucial to confirm the developed system’s usability, users’ safe and easy navigation and interactions with less workload, users’ accessibility, and realistic and practical functionality of the system. The usability of a system refers to how easy it is for users to learn, understand, and use the system to achieve their goals effectively and efficiently. Usability is an essential aspect of user experience design and development, as it directly impacts the satisfaction and productivity of users. The System Usability Scale (SUS), developed by Brooke (1996), is a robust and versatile tool for measuring the perceptions of usability (Bangor et al., 2008). It is a popular subjective measurement tool comprising a 10-item questionnaire with a five-point Likert scale, ranging from "strongly agree" to "strongly disagree." SUS can be useful for assessing the usability of the VR platform’s interface, controls, and overall experience. Participants rate their experience with a system under study on each item, and the results are combined to generate a single SUS score ranging from 0 to 100. A score of 68 is considered average, with higher scores indicating better usability (Sauro 2011). The SUS has been previously used to investigate different virtual reality systems (Webster and Dues 2017).

Workload refers to the available physical and cognitive resources required to perform or complete tasks. It can be affected by various factors, such as the job’s complexity, the amount of information presented, the level of experience and expertise of the individual, and the level of stress or fatigue. In the context of training and learning, the workload can have a significant impact on the effectiveness of the training or learning experience. If the workload is too high, individuals may become overwhelmed, leading to decreased performance and retention of information. On the other hand, if the workload is too low, individuals may become bored or disengaged, leading to decreased motivation and effort. To optimize workload during training and learning, it is essential to carefully design the learning materials and activities to ensure they are appropriately challenging for the intended users. This can involve breaking down complex concepts into smaller, more manageable pieces of information, providing clear and concise instructions, and incorporating interactive activities and assessments to help individuals actively engage with the material. Virtual training will likely support these requirements to optimize users’ workload in training. The NASA Task Load Index (NASA-TLX), developed by Hart (1986), is a subjective workload assessment tool used to measure a task or system's perceived mental, physical, and temporal demands. This tool has also been used to determine the perceived work effort inside a VR module in much past research (Dayarathna et al. 2020; Kournaditis, Chinello, and Venckute 2018; Armougum et al. 2019; Marucci et al. 2021).

To confirm users' safe exposure to virtual reality, researchers assess the simulation sickness induced in a participant from exposure to VR. One great tool to evaluate is the Simulation Sickness Questionnaire or SSQ (Kennedy et al. 1993). The questionnaire asks participants to rate sixteen symptoms of sickness from the simulation on a four-point scale (0–3). A total score of 5 or more indicates unsafe health conditions for VR exposure and does not allow users to continue using VR. This can be a limitation of VR training which can only be addressed by improving graphics quality and enhancing the realism of the immersion experience. However, safe VR exposure will always be limited, even with an optimized quality build. The VR platform’s realism can be tested by selecting important features used in the development. A realism score for these features can be collected on a 5-point Likert scale, with 1 being the least realistic and 5 being the most realistic. These scores can be used for the face validity of the VR platform by using an informal review of VR elements from non-experts. Face validity refers to the extent to which each selected feature is close to reality. Many past studies used face validity to confirm the realistic presentation of objects within a VR platform (Deb et al., 2017;Sethi et al.,2009).Carruth (2017) noted that the objective of VR is to facilitate the transfer of knowledge from VR to the real world. To evaluate the effectiveness of the virtual system, it is essential to expose novice trainees to the VR training, collect data on their performance, and transfer it to the real-world task to compare the real-world performance with those who only had traditional training. A great example would be the research conducted by Thomsen et al. (2017), where a high correlation between virtual reality and real-life performance for complex tasks like cataract surgery was observed. This research also emphasized forming groups with and without VR training to compare their performance on a virtual reality simulator and motion-tracking metrics from real-life cataract surgery.

This study used a virtual training platform for the compression molding process. Two operator groups were compared in their in-situ compression molding tasks: one group received virtual training while the other did not. Both subjective and objective measures were collected to investigate significant differences between the two groups’ performance and workload. The findings shed light on the potential of virtual training as a workforce development tool for improving worker performance and reducing their workload in real-world industrial training settings while effectively transferring knowledge from virtual tasks to real-world tasks.

Eight graduate students from the Human Engineering course were selected for an experimental study to investigate the effect of VR training and classroom lectures on trainees’ task performance and mental workload during training. The virtual training was conducted at the Human Factors Lab at The University of Texas at Arlington (UTA). The in-situ testing with the physical machine was performed at The University of Texas at Arlington Research Institute (UTARI).

The study was approved by the Institution of Research Board at UTA. The training was designed to satisfy all the basic needs of the practical operation. At first, the visual requirements for the VR system were identified, followed by the condition of types of interactions to complete the tasks. When all of them were incorporated in the final training, a familiarization module was built with the basic visuals and interaction systems.

3.1 Finalizing the List of Tasks

While producing a fiber-reinforced composite, a user must go through a list of tasks, as shown in Fig. 2. The researcher visited the actual compression molding facility at UTARI before any VR development to understand the entire process. However, a simplified version of the process was designed because of the need for more time and resources for VR development and to reduce the complexities of the task list (Fig. 3).

3.2 Choosing VR Development Method and Apparatus

After listing the tasks to be completed in the VR training, finalizing the platform where the VR would be developed and the apparatus to run the training was necessary. Unity was chosen as the VR development platform as it is free for non-commercial purposes and available online resources to develop games using modern virtual reality headsets. The apparatus chosen for this VR development was the HTC Vive Pro. This device has a two-controller system with a wide range of controlling options inside a virtual environment. It has a higher display resolution than the consumer-favorite low-cost and portable headset Meta Quest 2 and provides a larger field of view. HTC Vive system uses external base stations for tracking, which offers more precise and accurate tracking than its competitors.

3.3 Identifying the Visual Requirements

After visiting the real-world facility, the following visual requirements were found to be necessary to complete the virtual environment (Fig. 4): (1) An operation room where the machine is situated; (2) A safety station outside the preparation room so that the user can have them picked before they start any process; (3) A preparation room where the user prepares the raw materials to be processed in the machine; and (4) There are two tables in the preparation room to pick up objects to assemble on a third table – the primary workstation for preparing the lay-up. A developed version of this lab layout in Unity can be seen in Fig. 5.

3.4 Identifying the VR Interaction Requirements

Virtual Reality provides a range of interaction modes that enable users to engage with virtual environments and objects. For example, controller-based, index finger-based, gaze-based (hands-free), voice-based, or haptic-based interactions. These interaction modes can be combined to create more immersive and engaging experiences in VR. The interaction method for this study was chosen based on the type of tasks the user would have to go through to complete the training (Ely, Thomson, and Orlansky 1956). There were six types of interaction necessary to complete this VR training:

3.4.1 Toggling the Task List

The task list directs the user to the required tasks they need to complete for the training. The task list consisted of six tasks that were supposed to be completed by the user. A double press on the menu button of the left controller toggles on or off the list of tasks to be completed by the user. Once the user completes a task, it gets highlighted green in the list. The task list can be seen in Fig. 6.

3.4.2 Teleportation

The teleportation method would allow the user to move inside the virtual environment from one place to another. Twelve teleportation points in the module allowed access to machines, workstations, and safety stations, among other vital locations within the virtual lab (shown in Fig. 7)

3.4.3 Object Grab and Release

The training module demanded that the user be able to pick up and drop objects at defined workstations to ensure all the tasks were complete. A grabbing mechanism using the left controller was set up to enable the users to complete the required actions. The user would use the left controller grab button by pressing the controller using the wrist (Fig. 8). The object is released as soon as the user touches a destination point and releases the grab button with the object (Fig. 9).

3.4.4 Interactable Digital Screen

There were two instances where digital screens were used inside the virtual environment. The first one is the digital screen on the machine that allows the user to set up the recipe (operational parameters) for the manufacturing process. The second one is on a thermostat outside the preparation room, allowing the user to check the temperature inside the room. The interaction with the digital screen was possible using the ray-casting method installed in the right controller to point at a menu button and later click on it through the trigger operation (shown in Fig. 10).

3.4.5 Interactable Machine Buttons

The compression-molding machine in real life requires the user to complete the action of opening and closing the machine's clamp and door using push buttons. It was achieved by a slight push of the virtual button of the user on the push buttons. Similar push buttons on the machine body were simulated in the VR training so that the user could get an experience as close to reality as possible (shown in Fig. 11).

3.4.6 Familiarization Module

A familiarization module with all the basic functionalities of the VR training was built so that the users could get used to all types of interactions and required skill sets to complete the training. This module involved only a list of four tasks. The first task asked it to interact with a screen, which is required in the VR training to set the machine recipe (Fig. 12). The second task was to use a safety object (Fig. 13). The third task was to pick an object from a table and put it on top of another, which showed similar actions of preparing the lay-up in the preparation room in the main training module (Fig. 14). The final task was to push a button, which imitated pressing a button in a machine (Fig. 14)

3.5 Study Design

Graduate students were asked to participate in a theoretical lecture on manufacturing fiber-reinforced composites. Then they were randomly chosen to form into two groups where one group went through the VR training, while the other did not. Objective and subjective data were collected throughout the study to compare the impact of VR training between the two groups. Finally, both groups participated in conducting the process in real-world (Fig. 15).

3.5.1 Classroom Lecture

The graduate participants attended a lecture by a field expert on manufacturing fiber-reinforced composites. They were given a theoretical understanding of composites and how fiber-reinforced composites are formed using epoxy resins. In addition, a brief overview of these composites was shared with the students to enhance their understanding of their widespread applications in the industry. The lecture also included scientific information regarding the materials required for manufacturing. The layup and curing methods of the process were also discussed. Three segments of curing recipes are discussed in the lecture: (1) Ramp-up (increasing temperature and pressure); (2) Hold/ dwell; and (3) Ramp-down (decreasing temperature and pressure).

3.5.2 Group Formation

Students from a graduate course were chosen to be randomly placed into two groups (A and B) to participate in the experiment. Each group consisted of four participants. Both groups of students participated in the theoretical lecture session. After the lecture, participants of Group A were asked to participate in the VR training, whereas Group B was not. Both groups were asked to visit a compression molding laboratory and complete the machine operation steps under supervision. It is costly to conduct the actual operations for the participants, given that the machine is continuously used in a laboratory where participants' intervention would reduce the available time of the machine in the facility. Also, it is a very high tonnage (50 tons) compression molding machine. Therefore, any safety constraints violated can be injurious to the participants.

3.5.3 Virtual Reality Training

Only the Group A participants were eligible for the VR Training. At first, the participants were asked to share some demographic information in a survey (e.g., Age, Gender, familiarity with VR, etc.). Then they were asked to complete the Simulation Sickness Questionnaire (SSQ) (Kennedy et al. 1993) before they were introduced to the familiarization module. The SSQ was used three times during the VR training, as the quantitative result (below 5) was a factor in allowing them to complete the next VR activity. During the familiarization, they learned how to use the controller, headset, and interaction in the virtual environment. After that, they were given the SSQ again, followed by the final VR training. The SSQ was completed again for the last time after completing the VR training, which led them to complete a user experience survey that included system usability ratings and subjective data on the user experience (Fig. 16).

3.5.4 Real-world Exercise

Group A and B participants were invited for a hands-on experience to conduct the manufacturing process of fiber-reinforced composites at a production facility. First, they were given a small briefing on how to operate the process and control the machine. After that, they were asked to complete the process by themselves in the presence of an expert. The expert helped them but only intentionally interfered when necessary. After completing the task, they were asked to fill up a survey on their experience. The survey measured the workload and asked about their level of comfort and confidence. The questions differed from Group A to Group B – as some questions were added for the Group A participants to evaluate their opinions on the effectiveness of the VR training (Fig. 17).

3.6 Data Collection

Different types of objective and subjective measures were collected from virtual reality training and real-world exercises. The objective criteria are quantitative, and the subjective data is qualitative.

3.6.1 Objective Measures

The objective measures in this research refer to the quantitative methods of data collection that produce numerical data for statistical analysis. The objective measures used in this research are:

Time Measurement: The time taken to complete the operation on the physical machine for both groups was calculated. Afterward, a comparison of the time between the two groups was conducted.
Task Accuracy: The task accuracy/error rate was calculated for both groups at the real workstation. It was followed by a comparison of the accuracy between the two groups. User accuracy was calculated on a total of ten tasks performed, the first task involving three subtasks. Each task, if appropriately completed, was given a rating of 1, else 0 for a failure to complete the task. The tasks included (1) wearing Safety Gear, including picking up gloves, Hairnet, and Apron; (2) setting the recipe in the machine; (3) using mold plates; (4) using release films; (5) using peel plies; (6) properly using prepreg for layups; (7) placing lay-up in the machine; (8) closing the ramp of the machine; (9) closing the door of the machine; and (10) initiating recipe.

3.6.2 Subjective Measures

The subjective measures in this research are the methods of collecting data that rely on the participants’ personal opinions, experiences, or perspectives. The subjective data considered for the research are:

System Usability Score: The system usability score rating of the virtual environment. It is based on the users’ opinions, experiences, and feelings about the VR system.
The system usability score is based on ten questions asked to the user on a scale from 1 to 5 ranging from Strongly Disagree to Strongly Agree. The odd-numbered questions leaned towards the positive aspects of the system, and the even-numbered questions leaned towards the negative aspects of the system.
Face Validity: The participants were asked to rate some VR elements (Safety Equipment, Preparation Room, Carbon Fiber Laminates, Compression Molding Machine, Machine Screen, Recipe Menu, Machine Switches) on a scale of 5, with 5 being the highest level of realistic attributes the virtual item was able to show.
Confidence of the participants: A survey was conducted to gather feedback on the confidence level before and after the operation between the two groups.
Workload Measurement: Six workload dimensions were collected according to the NASA Task Load Index, which involves rating six workload dimensions on a scale from 0-100, where 0 represents "no workload" and 100 represents "maximum workload." The six dimensions are Mental Demand; Physical Demand; Temporal Demand; Performance; Effort; and Frustration Level. The workload was computed for both the VR training and physical application for both groups.
Feedback on the training - Feedback from both the groups receiving the VR training on their experience of the VR training was collected.

4.1 Objective Measures

4.1.1 Time Measurement

The average time to complete all tasks in the real-world exercise for the participants of Group A was 15.98 minutes, whereas the average time taken for the participants of Group B was 16.18 minutes. The time management results are presented in Table 1. The average time taken to complete the VR task was 9.35 minutes.

Table 1

Time taken for both groups to complete all the tasks in the real world.
Group	Minimum time to complete tasks (mins)	Maximum time to complete tasks (mins)	Mean (mins)	Variance (mins)
A (with VR training)	12.17	19.00	15.98	8.81
B (without VR training)	13.83	19.43	16.18	4.16

Based on the mean vs. variance analysis from Table 1, we can conclude that the virtual reality training might have helped Group A participants to complete the task slightly faster on average. However, there is a larger variability in Group A's completion times, which might indicate that the effectiveness of the virtual reality training varies among individuals.

4.1.2 Task Accuracy

Task Accuracy for all tasks completed in the real world was calculated for both groups of participants (shown in Table 2). Group A had an average accuracy of 96% for all the tasks completed. Group B had an average accuracy of 85% for the completed tasks. A breakdown of task-by-task accuracy for both Groups A and B is shown in Figs. 18 and 19.

Table 2

Overall Task Accuracy of Groups A and B
Group	Highest Individual Task Accuracy	Lowest Individual Task Accuracy	Average Task Accuracy
A	100%	83%	96%
B	92%	85%	85%

4.2 Subjective Measures

4.2.1 System Usability Scale

The System Usability Scale (SUS) score was calculated from survey data of Group A students on their experience with the VR training. Users were asked to rate 10 questions on a scale from 1 to 5, where 1 meant Strongly Disagree, and 5 meant Strongly Agree. The scores were adjusted by subtracting 1 from the response of the odd-numbered items, subtracting the user responses of the even number items from 5, and finally multiplying the scores by 2.5 (Sauro 2011). his converts the range of possible values from 0 to 100 instead of from 0 to 40. The System Usability Score, after necessary adjustment, came to a total sum of 90 (Table 3). A SUS score above 68 is considered above average (Sauro 2011). So, we can conclude that the SUS score observed supports the efficacy of the training platform.

Table 3

Scores from System Usability Scale (VR Training)
Question	Mean Score	Adjusted Score
I think that I would like to use this system frequently.	4.75	9.38
I found the system unnecessarily complex.	1.00	10.00
I thought the system was easy to use.	5.00	10.00
I would need the support of a technical person to be able to use this system.	2.00	7.50
I found the various functions in this system were well integrated.	4.75	9.38
I thought there was too much inconsistency in this system.	1.00	10.00
I would imagine that most people would learn to use this system very quickly.	4.25	8.13
I found the system very cumbersome to use.	2.25	6.88
I felt very confident using the system.	4.75	9.38
I needed to learn a lot of things before I could get going with this system.	1.25	9.38

4.2.2 Face Validity

The users were asked to rate the below VR element on a scale of 1 to 5, where 1 meant it was least realistic and 5 meant it was most realistic. It was seen that all the virtual environment elements had an average score of more than 4 which represents that the participants found the objects in the virtual environment very realistic (Fig. 20).

4.2.3 Confidence of the Participants

The confidence level was subjective data the participants were asked to state their level of confidence on a scale of 10 before and after the task completion. It can be observed that Group A participants had an average confidence level better than Group B students before starting the task, and after completing the task (shown in Fig. 21)

4.2.4 Workload in the VR Training

The NASA Task Load Index (TLX) method was used to assess the workload of Group A, participants completing VR training, on six workload dimensions on a scale of 0 to 100 for VR training (Fig. 22).

The study showed that the participants had a low level of Physical Demand (7.5) and Temporal Demand (1.25) while participating in the VR training. It shows that the task did not require much physical activity and was not time sensitive which was obvious due to the addition of teleportation to reduce movements and for fast pacing the manufacturing process. The mental demand score (33.75) means that the task required moderate levels of cognitive processing and mental effort. This could mean that the task was moderately complex and required some degree of concentration and focus. The mental workload can also be the result of exposure to VR training. A high score in performance (98.75) indicates that the task was well-defined, interactions and navigations were designed effectively, and the participants felt competent in their ability to perform the task. The effort score (8.75) suggests that the participants experienced a moderate level of effort during the task. This could be related to the mental effort required to complete the task, as seen in the higher mental demand score. Finally, the frustration score (1.25) suggests that the participants did not experience much frustration during the task and enjoyed the emerging technology-based learning.

4.2.5 Workload in the Real-World Task

The NASA Task Load Index (TLX) method was used to assess the workload of both groups on six workload dimensions on a scale of 0 to 100 for the real-world task. After calculating the workload, it is seen that, except for performance, for all dimensions of the workload, the participants of Group B had a higher workload in the real-world task (Fig. 23). A higher score in the performance sub-criteria for the Group A participants means that they probably felt higher confidence in performing the task very well. In addition, Group B participants stated that they had a higher level of frustration (16.25) than Group A participants (1.25).

Furthermore, a one-sided t-test was completed in the IBM SPSS Statistics software (version 29.0.1.0) on the six sub-criteria at a 90% confidence level to check with a null hypothesis (H0) that there is no significant difference between Group A and Group B in terms of their ratings for each sub-criterion and an alternative hypothesis (H1) that there is a significant difference between Group A and Group B in terms of their ratings for each sub-criterion. The results are shared below in Table 4:

Table 4

t-Test results for the workload
Sub criteria	p-value at 90% Confidence level	Significant difference observed?
Mental Demand	0.077	Yes
Physical Demand	0.167	No
Temporal Demand	0.043	Yes
Performance	0.095	Yes
Effort	0.024	Yes
Frustration	0.119	No

From the t-test results, it was observed that there are significant differences between Group A and Group B in terms of their ratings for mental demand, temporal demand, performance, and effort. However, there are no significant differences between the groups in terms of their ratings for physical demand and frustration. A low sample size can be the reason for this lack of statistical evidence regarding the difference between the two group’s workloads.

4.2.6 Workload Comparison of VR Task and Real-World Task

Here is a comparison between the VR Task and the real-world task workload of Group A participants:

It can be observed from Fig. 24 that the mental demand for the VR task (33.75) was higher than for the real-world task (17.50). A higher score usually means that the task requires significant cognitive processing, mental effort, attention, and concentration. At the same time, if we look into the workload data of Group B students, we can see that they have a workload score of 41.25, which is 136% more than the Group A students. So, it is possible that the Group A students only rated their real-world task mental demands lower as they have had a VR task advantage over the other group.

4.2.7 Simulation Sickness Questionnaire

For all the participants of the VR task, the Simulation Sickness Questionnaire was provided before the familiarization module, after the familiarization module, and after the VR training. For all SSQs, it was ensured that the cumulative score was below 5 following any VR activity. This result confirms that the developed training platform was appropriately designed with high-quality graphics, realistic immersion, interaction, and control activity along with adequate exposure time.

The study focuses on using VR training to enhance the learning experience of a complex real-world task involving physical interaction with materials and machines. The study tried to investigate if the addition of VR training with traditional training improves the task performance of the operator. It also evaluates the impact of VR training on the workload of a real-world task. Participants were divided into two groups: Group A had VR training added to the traditional training, and Group B only participated in the traditional training. Objective and subjective data were collected.

In the objective measures, time management and task accuracy data were collected. The time management data in Table 1 shows that Group A has taken a lower average time to complete all tasks than Group B. However, it is only 1.7% lower than the time taken by Group B which does not provide any evidence of a significant difference. There were reductions in task accuracy for some of the complex tasks performed by the group without VR training (Figs. 19 and 20). These tasks are recipe setting, peel ply use, ramp closure, and recipe initiation. Only the door closure task was more accurately conducted without VR.

System Usability Scale, NASA TLX, confidence response, and Simulation Sickness Questionnaire surveys were administered for subjective measures. A list of features was selected to be rated for the realism score to conduct face validity of the developed system. A high score of 90 in SUS shows the efficacy of the system in training students for composite manufacturing. Higher SUS score was also previously observed byDeb et al (2017) in past research where participants provided higher usability scores for an immersive and interactive platform-based system. Group A participants had an average confidence level better than group B, especially before starting the task. Once they finished the task in the real world, their confidence level was about the same. Which emphasized more assistance within the virtual environment to improve trainees’ learning experience.

Students going through the VR training experienced very low physical and temporal demands along with low frustration levels and effort. However, they have experienced a very high level of performance showing a high level of satisfaction from the VR training. It could be true that VR participants felt pressure to perform well in real-world tasks too. Face validity was conducted for the elements in the virtual environment, which showed that the participants found the virtual environment very realistic. One participant commented that basic VR knowledge is sufficient for first-time users, while another suggested using a wireless headset because they sometimes found the wired headset uncomfortable. All participants agreed that it was easy to understand the tasks in the VR because of the friendly and assistive task list.

It was observed that the VR trainees (Group A) performed significantly better than Group B in accurately completing the task. This positive effect of VR training in completing tasks with lower errors was seen in research conducted by Seymour et al. (2002) It was also found that the Group A participants went through a lower workload in the real-world task than the participants of Group B. It was also found that the Group A students had a significantly reduced mental demand to complete the real-world task because of the VR training. The frustration level was much higher for Group B participants in the real-world task than for Group A students. A difference in the task completion time of the real-world task could not be found. However, the average time taken to complete the task in VR was very low compared to the real-world task. There were no issues with simulation sickness of the virtual platform confirming participants' safe exposure to the VR training.

Some limitations of this research include the experiment being designed with only eight participants forming into two different groups due to the cost and safety constraints of the real-world task. The VR training also did not have any audible instructions or sounds corresponding to the virtual environment. The real-world task was ended after the recipe initiation of the machine as it would require a long time for the participants to wait for the curing to complete and take out the processed part. However, utilizing the fast simulation in VR, it was possible to let the user continue at least one more step (to open the clamp and door of the machine) after the curing is complete.

The research helped understand the effect of additional VR training along with traditional training for complex real-world tasks. It showed that VR training effectively prepares the workforce for composite manufacturing. A person trained in VR is more confident and less frustrated while doing a real-world task for the first time. VR training is cost-efficient and can be conducted multiple times by the same user without depreciating a real-world machine for training purposes.

Using fast simulation and extended interactive techniques, the VR training can be more realistic, and the whole training can be more accurate to the real-world task. The VR training was made to be fail-proof, which meant that if the users' activities were not completed in the task list sequentially, it wouldn’t let the user complete the next step. While developing the training, the expectations were to include all steps of the process in VR, but in reality, it was only possible to add some due to a limited workforce. An upgraded version of the training can be built with the users allowed to make mistakes so that they can check the outcome of wrong decisions made in the game. In the future, the inclusion of Augmented Reality along with VR to train on such manufacturing tasks can be tested to find improvement in task performance. A comparison of how investing in VR training saves time and cost can be made to make it a more viable case for industrial use. Future research on the effectiveness of VR training already implemented in the industry can help researchers find new and modified methods of the applications of VR to help spread the next-generation technology of workforce preparation.

Author Contribution

Taufiq Rahman is the lead researcher and writer of this manuscript.Minhazur Rahman is a co-researcher who has actively organized the trials and collected data required for this project.Rassel Raihan is an advisor and has trained the trial participants on the necessary knowledge required to participate in the trials.Shuchisnigdha Deb is the lead supervisor of this research - who has guided the team on how to complete the research, analyze data and improve the quality of the study.

Abliz D, Duan Y, Steuernagel L, Xie L, Li D, and Gerhard Ziegmann (2013) Curing Methods for Advanced Polymer Composites -A Review. Polym Polym Compos 21(6):341–348. https://doi.org/10.1177/096739111302100602
Angelov V, Petkov E, Shipkovenski G (2020) and Teodor Kalushkov. Modern Virtual Reality Headsets. HORA 2020–2nd International Congress on Human-Computer Interaction, Optimization and Robotic Applications, Proceedings, June. https://doi.org/10.1109/HORA49412.2020.9152604
Armougum A, Orriols E, Gaston-Bellegarde A, Joie La C, Marle, Piolino P (2019) Virtual Reality: A New Method to Investigate Cognitive Load during Navigation. J Environ Psychol 65(October):101338. https://doi.org/10.1016/J.JENVP.2019.101338
Bangor A, Kortum PT, Miller JT (2008) An Empirical Evaluation of the System Usability Scale. Https://Doi.Org/10.1080/10447310802205776 24 (6): 574–94. https://doi.org/10.1080/10447310802205776
Bell JT, Scott Fogler H (1995) The Investigation and Application of Virtual Reality as an Educational Tool.
Bhatnagar N, Asija N (2016) Durability of High-Performance Ballistic Composites. Lightweight Ballistic Composites: Military and Law-Enforcement Applications: Second Edition. May, pp 231–283. https://doi.org/10.1016/B978-0-08-100406-7.00008-8.
Bozgeyikli E, Raij A, Katkoori S (2016) and Rajiv Dubey. Point & Teleport Locomotion Technique for Virtual Reality. CHI PLAY 2016 - Proceedings of the 2016 Annual Symposium on Computer-Human Interaction in Play, October, 205–16. https://doi.org/10.1145/2967934.2968105
Brooke J, June (1996) 207–212. https://doi.org/10.1201/9781498710411-35
Carruth DW (2017) Virtual Reality for Education and Workforce Training. ICETA 2017–15th IEEE International Conference on Emerging ELearning Technologies and Applications, Proceedings, November. https://doi.org/10.1109/ICETA.2017.8102472
Chénéchal M, Le, Chatel-Goldman J (2018) HTC Vive Pro Time Performance Benchmark for Scientific Research. November. https://hal.science/hal-01934741
Choi S, Jung K, and Sang Do Noh (2015) Virtual Reality Applications in Manufacturing Industries: Past Research, Present Findings, and Future Directions 23. 140–63. https://doi.org/10.1177/1063293X14568814
Coburn JQ, Freeman I, Salmon JL (2017) A Review of the Capabilities of Current Low-Cost Virtual Reality Technology and Its Potential to Enhance the Design Process. J Comput Inf Sci Eng 17(3). https://doi.org/10.1115/1.4036921/370980
Dai S (2022) Design and Implementation of Unity3D Virtual Reality-Based Simulation System for Dance Studies. Proceedings – 2022 International Conference on Data Analytics, Computing and Artificial Intelligence, ICDACAI 2022, 376–79. https://doi.org/10.1109/ICDACAI57211.2022.00080
Dayarathna VL, Karam S, Jaradat R, Hamilton MA, Nagahi M, Joshi S, Ma J, Ashour O, and Bouteina Driouche (2020) Assessment of the Efficacy and Effectiveness of Virtual Reality Teaching Module: A Gender-Based Comparison. Int J Eng Educ 36(6):1938–1955. https://pennstate.pure.elsevier.com/en/publications/assessment-of-the-efficacy-and-effectiveness-of-virtual-reality-t
Deb S, Carruth DW, Sween R, Strawderman L, and Teena M. Garrison (2017a) Efficacy of Virtual Reality in Pedestrian Safety Research. Appl Ergon 65:449–460. https://doi.org/10.1016/j.apergo.2017.03.007
Deb S, Carruth DW, Sween R, Strawderman L, and Teena M. Garrison (2017b) Efficacy of Virtual Reality in Pedestrian Safety Research. Appl Ergon 65:449–460. https://doi.org/10.1016/J.APERGO.2017.03.007
Doolani S, Wessels C, Kanal V, Sevastopoulos C, Jaiswal A, Nambiappan H, Makedon F (2020) A Review of Extended Reality (XR) Technologies for Manufacturing Training. Technol 2020 8(4):77. https://doi.org/10.3390/TECHNOLOGIES8040077
Ely, Jerome H, Robert M, Thomson (1956) and Jesse Orlansky. Chapter VI of the Joint Services Human Engineering Guide to Equipment Design. In Design of Controls
Erden Seçkin (2017) and Kingsley Ho. Fiber Reinforced Composites. Fiber Technology for Fiber-Reinforced Composites, January, 51–79. https://doi.org/10.1016/B978-0-08-101871-2.00003-5
Erogonomiesite.be. n.d (2023) Nasa Task Load Index. Accessed May 2, https://www.ergonomiesite.be/documenten/risicoanalyse/NASA-TLX.pdf
Fang H, Xiru Y, Kun H (2021) Li Wenxin, and Tian Haodong. Design and Application of VR Lab Based on Unity. Journal of Physics: Conference Series 1982 (1): 012167. https://doi.org/10.1088/1742-6596/1982/1/012167
Fatima A, Bukhari F (2022) and Waheed Iqbal. Development of Low-Cost Surgical Simulator for Neuroendoscopy Using Unity3D and HTC VIVE. 2022 16th International Conference on Open Source Systems and Technologies, ICOSST 2022 - Proceedings. https://doi.org/10.1109/ICOSST57195.2022.10016865
Fiber-Reinforced Composites Market – Global Industry Trends and Forecast to 2028 | Data Bridge Market Research. n.d. Accessed April 17 (2023) https://www.databridgemarketresearch.com/reports/global-fiber-reinforced-composites-market
Fortune BI (2019) Virtual Reality (VR) in Education Market Size, Research, Growth Report 2026. September 2019. https://www.fortunebusinessinsights.com/industry-reports/virtual-reality-in-education-market-101696
Gabajová G, Krajčovič M, Matys Marián, and Beáta Furmannová (2021) Designing Virtual Workplace Using Unity 3D Game Engine. Int Sci J about Int Sci J about Int Sci J about Technol Technol Technol 7:35–39. https://doi.org/10.22306/atec.v7i1.101
Global MI (2022) Carbon Fibre Composites Market Share and Statistics – 2030. July 2022. https://www.gminsights.com/industry-analysis/carbon-fibre-composites-market
Hart SG (1986) TASK LOAD INDEX (NASA-TLX) v 1.0. https://ntrs.nasa.gov/api/citations/20000021488/downloads/20000021488.pdf
Ikbal M, Sadiq V, Ramadoss, and Matteo Zoppi (2021) Dynamic Pose Tracking Performance Evaluation of HTC Vive Virtual Reality System. IEEE Access 9:3798–3815. https://doi.org/10.1109/ACCESS.2020.3047698
Ip P (2022) Virtual Reality in Education: How Schools Are Using VR – 42West. Adorama.Com. December 7, 2022. https://www.adorama.com/alc/virtual-reality-in-education/
Jimeno A, and Alberto Puerta (2007) State of the Art of the Virtual Reality Applied to Design and Manufacturing Processes. Int J Adv Manuf Technol 33(9–10):866–874. https://doi.org/10.1007/S00170-006-0534-2/METRICS
Kennedy RS, Norman E, Lane, Kevin S, Berbaum, Lilienthal MG (1993) Simulator Sickness Questionnaire: An Enhanced Method for Quantifying Simulator Sickness. Int J Aviat Psychol 3(3):203–220. https://doi.org/10.1207/S15327108IJAP0303_3
Kournaditis K, Chinello F (2018) and Sarune Venckute. Design of a Virtual Reality and Haptic Setup Linking Arousals to Training Scenarios: A Preliminary Stage. 25th IEEE Conference on Virtual Reality and 3D User Interfaces, VR 2018 - Proceedings, August, 613–14. https://doi.org/10.1109/VR.2018.8446528
Lee SY, Seo J, Kim G, Lee C-MPSY, Kim GJ, Park C-M (2003) and Sangyoon Lee. Evaluation of Pointing Techniques for Ray Casting Selection in Virtual Environments. Https://Doi.Org/10.1117/12.497665 4756 (1): 38–44. https://doi.org/10.1117/12.497665
Liu G (2019) and Ji Wu. Design and Implementation of Virtual Campus Roaming System Based on Unity3d. Proceedings – 2019 International Conference on Machine Learning, Big Data and Business Intelligence, MLBDBI 2019, November, 147–50. https://doi.org/10.1109/MLBDBI48998.2019.00034
Luckett E (2018) A Quantitative Evaluation of the HTC Vive for Virtual Reality Research. Honors Theses, January. https://egrove.olemiss.edu/hon_thesis/331
Marsh H, Francisco R-R (2006) Production and Reference Material. Activated Carbon 454–508. https://doi.org/10.1016/B978-008044463-5/50023-6
Marucci M, Flumeri GD, Borghini G, Sciaraffa N, Scandola M, Pavone F, Babiloni F, Betti V, Aricò P (2021) The Impact of Multisensory Integration and Perceptual Load in Virtual Reality Settings on Performance, Workload and Presence. https://doi.org/10.1038/s41598-021-84196-8
Matsas E, and George Christopher Vosniakos (2017) Design of a Virtual Reality Training System for Human–Robot Collaboration in Manufacturing Tasks. Int J Interact Des Manuf 11(2):139–153. https://doi.org/10.1007/S12008-015-0259-2/TABLES/1
Matsas E, Vosniakos GC, and Dimitris Batras (2018) Prototyping Proactive and Adaptive Techniques for Human-Robot Collaboration in Manufacturing Using Virtual Reality. Robot Comput Integr Manuf 50(April):168–180. https://doi.org/10.1016/J.RCIM.2017.09.005
Moghadam K, Banigan C, and Eric D. Ragan (2020) Scene Transitions and Teleportation in Virtual Reality and the Implications for Spatial Awareness and Sickness. IEEE Trans Vis Comput Graph 26(6):2273–2287. https://doi.org/10.1109/TVCG.2018.2884468
Monetti FM, de Giorgio A, Yu H, Maffei A, Romero M (2022) An Experimental Study of the Impact of Virtual Reality Training on Manufacturing Operators on Industrial Robotic Tasks. Procedia CIRP 106:33–38. https://doi.org/10.1016/J.PROCIR.2022.02.151
Niehorster DC, Li, Li, and Markus Lappe (2017) The Accuracy and Precision of Position and Orientation Tracking in the HTC Vive Virtual Reality System for Scientific Research. I-Perception 8(3). https://doi.org/10.1177/2041669517708205
Njuguna J, Krzysztof Pielichowski, and Jiying Fan (2012) Polymer Nanocomposites for Aerospace Applications. Adv Polym Nanocomposites: Types Appl 472–539. https://doi.org/10.1533/9780857096241.3.472
Prashanth S, Subbaya Km, Nithin K, and Sachhidananda S (2017) Fiber Reinforced Composites -A Review. J Mater Sci Eng 6:341. https://doi.org/10.4172/2169-0022.1000341
Prithul A, Adhanom IB, and Eelke Folmer (2021) Teleportation in Virtual Reality; A Mini-Review. Front Virtual Real 2:138. https://doi.org/10.3389/FRVIR.2021.730792/BIBTEX
Quevedo WX, Jorge S, Sánchez O, Arteaga ÁlvarezV, Marcelo VíctorD, Zambrano CR, Sánchez, Andaluz VíctorH (2017) Virtual Reality System for Training in Automotive Mechanics. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 10324 LNCS: 185–98. https://doi.org/10.1007/978-3-319-60922-5_14/FIGURES/12
Rahman T, Santiago BD, Rimu A, Deb S, Pande A, Islam M (2022) A Virtual Reality Program to Improve Child Pedestrians’ Safety at Street-Crossing Scenarios. Hum Factors Syst Interact 52. https://doi.org/10.54941/AHFE1002150
Rajak D, Kumar DD, Pagar PL, Menezes, and Emanoil Linul (2019) Fiber-Reinforced Polymer Composites: Manufacturing, Properties, and Applications. Polymers 11(10). https://doi.org/10.3390/POLYM11101667
Rawi S, Mohd S, Alias, and Siti Aishah Wahid (2020) Design and Development Mini Compression Molding for Teaching and Learning. Educ Social Sci Rev 1(2):48. https://doi.org/10.29210/07ESSR57300
Renner A, Holub J, Sridhar S, Evans G, Winer E (2016) A Virtual Reality Application for Additive Manufacturing Process Training. Proceedings of the ASME Design Engineering Technical Conference 1A-2015 (January). https://doi.org/10.1115/DETC2015-47807
Roldán J, Jesús E, Crespo (2019) Andrés Martín-Barrio, Elena Peña-Tapia, and Antonio Barrientos. A Training System for Industry 4.0 Operators in Complex Assemblies Based on Virtual Reality and Process Mining. Robotics and Computer-Integrated Manufacturing 59 (October): 305–16. https://doi.org/10.1016/J.RCIM.2019.05.004
Salah B, Abidi MH, Mian SH, Krid M, Alkhalefah H, and Ali Abdo (2019) Virtual Reality-Based Engineering Education to Enhance Manufacturing Sustainability in Industry 4.0. Sustain 2019 11(5):1477. https://doi.org/10.3390/SU11051477
Sauro J (2011) Measuring Usability with the System Usability Scale (SUS) – MeasuringU. 2011. https://measuringu.com/sus/
Sethi AS, William J, Peine Y, Mohammadi, and Chandru P. Sundaram (2009) Validation of a Novel Virtual Reality Robotic Simulator. Https://Home Liebertpub Com/End 23(3):503–508. https://doi.org/10.1089/END.2008.0250
Seymour NE, Gallagher AG, Roman SA, O’Brien MK, Bansal VK, Andersen DK, Satava RM et al (2002) Virtual Reality Training Improves Operating Room Performance: Results of a Randomized, Double-Blinded Study. Ann Surg 236(4):458. https://doi.org/10.1097/00000658-200210000-00008
Sloan J (2016) Materials & Processes: Composites Fibers and Resins. CompositesWorld. 2016. https://www.compositesworld.com/articles/composites-101-fibers-and-resins
Stelzer PS, Cakmak U, Eisner L, Doppelbauer LK, Kállai I, Schweizer G, Prammer HK, and Zoltan Major (2022) Experimental Feasibility and Environmental Impacts of Compression Molded Discontinuous Carbon Fiber Composites with Opportunities for Circular Economy. Compos Part B: Eng 234(April):109638. https://doi.org/10.1016/J.COMPOSITESB.2022.109638
Tapan Bhatt, Alpa PP, Gohil (2018) and Vijaykumar Chaudhary. Primary Manufacturing Processes for Fiber Primary Manufacturing Processes for Fiber Reinforced Composites: History, Development & Future Research Trends. https://doi.org/10.1088/1757-899X/330/1/012107
Thomsen AS, Skou P, Smith Y, Subhi Mla, Cour L, Tang GM, Saleh, and Lars Konge (2017) High Correlation between Performance on a Virtual-Reality Simulator and Real-Life Cataract Surgery. Acta Ophthalmol 95(3):307–311. https://doi.org/10.1111/AOS.13275
TORAY (2020) 2510 Prepreg System Data Sheet PRODUCT NOTES. www.toraycma.com
Webster R, Dues JF (2017) System Usability Scale (SUS): Oculus Rift® DK2 and Samsung Gear VR®. ASEE Annual Conference and Exposition, Conference Proceedings 2017-June (June). https://doi.org/10.18260/1-2--28899
Yallew T, Berhanu E, Kassegn S, Aregawi, and Alula Gebresias (2020) Study on Effect of Process Parameters on Tensile Properties of Compression Molded Natural Fiber Reinforced Polymer Composites. SN Appl Sci 2(3):1–8. https://doi.org/10.1007/S42452-020-2101-0/FIGURES/6

No competing interests reported.

Appendixs.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Design and Development of a Virtual Reality Training Platform for Fiber-Reinforced Composite Manufacturing

Status:

Version 1

Abstract

Figures

1. Introduction

2. Literature Review

2.1 Fiber-Reinforced Composites in Industrial Applications

2.2 Manufacturing Technique for Carbon Fiber Composite

2.3 A Brief History of VR Training in Manufacturing

2.4 Effective VR System Design

2.5 Evaluation of Virtual Systems

3. Methodology

3.1 Finalizing the List of Tasks

3.2 Choosing VR Development Method and Apparatus

3.3 Identifying the Visual Requirements

3.4 Identifying the VR Interaction Requirements

3.4.1 Toggling the Task List

3.4.2 Teleportation

3.4.3 Object Grab and Release

3.4.4 Interactable Digital Screen

3.4.5 Interactable Machine Buttons

3.4.6 Familiarization Module

3.5 Study Design

3.5.1 Classroom Lecture

3.5.2 Group Formation

3.5.3 Virtual Reality Training

3.5.4 Real-world Exercise

3.6 Data Collection

3.6.1 Objective Measures

3.6.2 Subjective Measures

4. Results

4.1 Objective Measures

4.1.1 Time Measurement

4.1.2 Task Accuracy

4.2 Subjective Measures

4.2.1 System Usability Scale

4.2.2 Face Validity

4.2.3 Confidence of the Participants

4.2.4 Workload in the VR Training

4.2.5 Workload in the Real-World Task

4.2.6 Workload Comparison of VR Task and Real-World Task

4.2.7 Simulation Sickness Questionnaire

5. Discussion and Limitations

6. Conclusion and Future Work

Declarations

Author Contribution

References

Additional Declarations

Supplementary Files

Status:

Version 1