Virtual Reality in Chemical and Biochemical Engineering Education and Training

With the advent of digitalization and industry 4.0, education in chemical and biochemical engineering has undergone significant revamping over the last two decades. However, undergraduate students sometimes do lack industrial exposure and are unable to visualise the complexity of actual process plants. Thereby, students might graduate without adequate professional hands-on experience. Similarly, in the process industry, operator training-simulators are widely used for the training of new and skilled operators. However, conventional training-simulators often fail to simulate reality and do not provide the user with the opportunity to experience unexpected and hazardous scenarios. In these regards, virtual reality appears to be a promising technology that can cater to the needs of both academia and industry. This paper discusses the opportunities and challenges for the incorporation of virtual reality into chemical and biochemical engineering education with an emphasis on the fundamental areas of technology, pedagogy and socio-economics. The paper emphasises the need for augmenting virtual reality interfaces with mathematical models to develop advanced immersive learning applications. Further, the paper stresses upon the need for novel educational impact assessment methodologies for the evaluation of virtual-reality-based learning. Finally, an ongoing case study application is presented to briefly discuss the social and economic implications, and to identify the bottlenecks involved in the adoption of virtual reality tools across chemical and biochemical engineering education.


Introduction
With rapid advances in technology, digital solutions are being widely embraced for teaching and for learning purposes. This includes the use of online learning platforms, interactive digital applications, and collaborative environments in education. Furthermore, the easy availability of portable digital devices including notebooks and smartphones fostered the digital embracement for education. Recently, immersive virtual environments are also being considered as a viable option for education and training owing to the availability, affordability, and rapid advancements in these technologies .
In academia, growing student numbers combined with the limited availability of physical resources is leading educational institutions to explore novel teaching alternatives that are, primarily, costeffective in nature. Moreover, force-majeure situations such as the COVID-19 pandemic advocates for more educational robustness and distance learning options. Similarly, in industry, with J o u r n a l P r e -p r o o f 5 increased automation and complexity of modern process plants, advanced training methods are becoming more desirable for the effective training of operators. In this regard, three-dimensional (3D) virtual reality environments, similar to those being used in the gaming and film industry, can be effectively utilised to generate learning environments. Even though virtual reality is not a new technology, recent advancements have aroused increased interest among scholars . With the arrival of low-budget head-mounted displays (HMDs), immersive virtual environments are becoming more accessible for the purpose of education .
Virtual reality (VR) can be defined as an interactive, three-dimensional (3D) computer simulation that seeks to immerse the user completely in an artificial environment . In VR, a conscious effort is made to convince the user of being physically present in a virtual world . Such perceptions are created with the help of advanced hardware components including HMDs and sensory devices .
VR, and other mixed reality interfaces, provide an opportunity to deliver educational content that might not be easily possible when using traditional lecture-based education . To date, VR technology has been widely deployed for developing visual training simulators for a wide range of industries, including virtual reality-based flight simulators, surgery-training simulators, and driving simulators . developed as an open-source software and was designed to provide a basic understanding of fermentation technology. The authors reported that more than 90% of the students found the fermentation laboratory effective for performing hands-on experiments virtually. de las Heras and co-workers evaluated the learning perception of undergraduate chemical and biochemical engineering students after replacing a physical fermentation laboratory with a commercial virtual laboratory. The study was performed at the University of Auckland in collaboration with the Technical University of Denmark. In their study a commercially available fermentation virtual laboratory provided by Labster, a web-based virtual laboratory, was used .
Moving beyond small-scale virtual laboratories, there are a handful of studies involving industrialscale plants for education in literature . Even though there has been significant developments in the education sector, the undergraduate chemical engineering students are mostly unable to visualise the scale and complexity of actual process plants . Due to logistic constraints and associated financial expenses, it becomes difficult to conduct plant tours and industrial field trips for students on a regular basis. Moreover, students are not allowed to operate process equipment in a real plant. Therefore, undergraduate chemical engineering students often do not get the necessary exposure and the opportunity to apply process theory to real-world problems. Similarly, in the process industry, operator training simulators (OTS) are necessary for the training of skilled operators. However, the conventional training simulators fail to capture the realism and do not give operators the actual experience of various unexpected and hazardous scenarios .
VR seems to be a potential technology for chemical engineering education that can cater to the parallel needs of both university and industry. VR can be used to develop industrial-scale virtual J o u r n a l P r e -p r o o f 7 practice platforms that would aid in improving the overall understanding of process plants . In addition, VR can be used to replicate potential scenarios including start-up, emergency shut down, unsafe and hazardous scenarios, which can be visualised and replayed multiple times in safe, guided, and cost-effective environments .
Even though VR technology has been utilised for education over the past decade, a comprehensive review on its application for chemical engineering education does not exist in the literature. In the case of academia, Bell and Fogler discussed the status and research prospects for virtual reality in chemical engineering education. This is the only literature overview that has been carried out in this domain to date. In the case of process industry, Patle and co-workers extensively reviewed the contributions of VR for OTS from the year 2000 to mid-2017. However, the authors did not discuss the wider technological/ pedagogical challenges and overlooked the various mathematical models that may be present behind VR-based systems.
In this paper, a review is carried out on using VR for chemical engineering education, with a focus on applications in academia and in process industry. The paper will discuss the opportunities and challenges in the utilisation of VR for education with an emphasis on the key areas of technology, pedagogy and socio-economics. The paper will review the existing practices of educational impact evaluation for virtual reality in the chemical engineering industry. Further, it will discuss the wider pedagogical implications and the prospects for educational research in this domain. It is to be noted that the review will be applied to virtual-reality-based case studies, and not virtual laboratories, in light of the overall motivation of this work.
In terms of the structure of this paper, section 2 reviews the literature, which includes a review on virtual reality for chemical engineering academia and for process industry training. Thereafter, section 3 discusses technological opportunities, section 4 discusses wider pedagogical J o u r n a l P r e -p r o o f 8 implications, whilst, section 5 addresses social and economic implications, for VR in education and training respectively. Finally, section 6 presents the conclusions and the directions for future research.

State of the art
In this section, an exhaustive literature search is done to review the previous work and to identify research prospects for chemical engineering education and training. Google scholar and Scopus database were employed with the keywords "virtual reality", "immersive environment", and "head-mounted display" to analyse the current state of the art studies reported in the literature. In addition, the keywords "chemical engineering", "process systems engineering" and "unit operations" were used to screen the studies specific to the chemical engineering domain. Original research articles were reviewed without any particular emphasis placed on the publication timeline.

Virtual reality for chemical engineering in academia
Various studies can be found in the literature that have made efforts to bring real process plants to the university setting with the help of virtual reality technology. Some of these studies are reviewed in the forthcoming text.
The first known application of virtual reality for chemical engineering education was that developed by Bell and Fogler in the department of chemical engineering at the University of Michigan in Ann Arbor. They developed a virtual reality application, Vicher (Virtual Chemical Reaction Module), for chemical reaction kinetics and reactor design education at the undergraduate level. The module consisted of a welcome centre, two reactor rooms, and two microscopic exploration areas. The microscopic areas allowed the student users to observe the catalytic reaction mechanisms at a molecular scale.
J o u r n a l P r e -p r o o f 9 Norton and co-workers developed a virtual immersive environment to cater to the educational needs of both industry and university. They developed an interface based on spherical imagery of a real operating plant that was integrated with interactive learning activities. The interface was based on a Crude Distillation Unit (CDU) of BP's Bulwer Island Refinery in Brisbane, Australia.
The immersive platform was developed by using high resolution industrial photography.
Herritsch developed a desktop-based virtual reality interface in the department of chemical and process engineering at the University of Canterbury, New Zealand. The interface was based on the skim milk powder processing facility of the Fonterra Co-operative Group, New Zealand. The application integrated and presented a browser-based 360-degree panoramic view, process flow diagram, and a 3D map of the milk powder plant. The VR interface was developed in Adobe Flash Builder, which can be easily used by the student users. Further, Rahim and co-workers assessed the usability of the same desktop-based interface by recruiting eight undergraduate students for a study. The authors report that the student users were satisfied with the virtual reality application since most of them found it really easy to learn and navigate. It is worth mentioning that during the recent times, low-budget VR headsets for mobile phones, such as Google cardboard, enables several users to experience similar browser-based virtual environments at the same time .
Schofield developed a software, ViRILE (Virtual Reality Interactive Learning Environment), for undergraduate chemical engineering education. The software simulated the operation of a polymerisation plant that consisted of a reactor section and three distillation columns. The interface was developed based on process plant simulation data that was generated using steady state flow sheeting software (i.e. a commercial process simulator) and C++ programming. However, in this study, the authors do not discuss the educational impact of the proposed software on the student community.

J o u r n a l P r e -p r o o f
Ouyang and co-workers developed a virtual practice platform that included virtual section scenes of a chemical plant. The user has options to move around the plant, and to understand the structure and operation of various process equipment and pipelines. Further, they also evaluated the performance of the virtual platform by gathering feedback from the student users in the form of a questionnaire.
Pirola and co-workers developed a virtual immersive crude distillation unit (CDU) for undergraduate classroom education. The immersive platform allows the students to take a virtual tour of the plant and to get involved in time-bound interactive virtual exercises. Further, they also evaluated the educational impact of the project by conducting a test involving 22 questions on the theory of the CDU. The participants included 34 Bachelor degree students from the discipline of Industrial Chemistry at the University of Milan, Italy. The test was conducted both before and after performing the virtual exercises by the users. The authors report that the overall score obtained by the students doubled in the latter case followed by a strong positive feedback about the virtual learning platform.
The above studies show that virtual reality can be used to develop immersive learning environments for chemical engineering education. However, most of the studies do not discuss the educational impact of virtual reality on the undergraduate student community. The existing case studies only discuss the short-term benefits and ignore the wider implications. Therefore, novel methodologies are necessary for educational impact evaluation.
The software technologies that were employed for interface development in the listed case studies are summarised in Table 1.

Virtual reality for operator training in chemical industries
J o u r n a l P r e -p r o o f 11 The role played by operators in chemical industries is becoming increasing challenging due to increased automation in modern plants . In this regard, OTS are widely used for training new operators and for refreshing the skills of existing operators . Such training systems aid in enhancing the knowledge and the experience of industrial operators. However, the traditional twodimensional OTS fail to provide a sense of realism to their users.
The virtual reality-based OTS supports the effective training of operators in a safe and supervised environment. Such systems can simulate demanding operating conditions, thereby training the operators to perform the corresponding operations at the plant. This includes exposure to noise disturbances, improper lighting conditions, and risky situations without jeopardising the plant operators . Moreover, such training simulators can be used to replicate various rarely-occurring scenarios including start-ups, shutdowns, and accidents . In addition, virtual reality gives an opportunity to experiment coordination and communication that might involve a team of operators for task accomplishment . Such training interfaces can also be developed to be multilingual, and can be utilised to train process operators in their preferred language .
Various studies can be found in the literature that have examined the role of immersive virtual reality systems for operator training. Some of these studies are reviewed in the forthcoming text.
Chen and Zhou developed a virtual reality-based pipeline simulation system for chemical process training. They proposed a data-driven simulation framework for the pipeline system. In addition, they also discussed the process path calculation method for a set of pipelines involved in the transportation of fluids. This system enables the process operators to visualise the process path update in a real-time virtual environment and enhances the training experience by allowing the operators to visualise the effects of fluid flow in addition to the 3D virtual view of the pipeline system.

J o u r n a l P r e -p r o o f
Manca and co-workers proposed the integration of a dynamic process simulator and a dynamic accident simulator within an immersive virtual environment for operator training. The authors applied the proposed framework to a case study involving the process of hydrodealkylation of toluene to benzene. They simulated an accident caused by a hole in the pipe that connected the toluene storage tank to the reaction section. The study demonstrated the effectiveness of the dynamic interface for hazard identification, education and training of industrial operators.
Manca and co-workers proposed an automated assessment procedure for industrial operator training in a 3D virtual environment. It involves an automated computer algorithm which saves every result from a performance assessment module that dynamically tracks the operator. The authors also illustrated the approach with the help of a case study of a real polymerisation plant.
The automated procedure can be used as a feedback mechanism to track the progress during operator training.
Luo and co-workers proposed an integrated simulation framework based on virtual reality and dynamic process modelling. They developed a 3D model of an existing ethylene plant that could interact with the dynamic model and the associated control system. The interface allows the operators to walk through the virtual plant and to monitor the working of the units. Moreover, it facilitates operator training on alarms, controls and graphics that are similar to the actual chemical process plant.
These studies highlight that immersive virtual reality environments allow process operators to experience the actual feel of the plant conditions. Thus, the operators gain the necessary training to prevent accidents and to handle emergency situations in an effective manner. However, similar to the case of virtual reality in academia, the existing case studies do not propose a training assessment methodology to evaluate the operators. A standardised framework is necessary to J o u r n a l P r e -p r o o f 13 evaluate the level of learning acquired by the operators. An overview of the existing case studies for operator training in chemical industries is presented in Table 2.
Following the literature review, opportunities and shortcomings have been identified in three fundamental areas for chemical engineering education. This includes the spheres of technology, education and socio-economics as depicted in Fig. 1.
In the technology sphere, virtual reality is primarily used in chemical engineering education as a visualisation tool for viewing 3D models of process plants. However, it can be augmented with a set of comprehensive mathematical models for advanced applications in an immersive environment. At the same time, the technological readiness of undergraduate students to embrace virtual reality tools for education is an important factor .
In the education sphere, the literature exposes two key issues. Firstly, it identifies the need to quantify the impact of virtual reality tools for chemical engineering education. Secondly, the literature draws attention to the purposeful application of virtual reality tools for specific groups of learners with different learning goals.
Finally, the limited number of examples from the literature suggests that there is a bottleneck in the adoption of virtual reality tools across chemical engineering education. The associated challenges are discussed under the wider social and economic implications. This paper will address each of the above spheres in the sections that follow.

Technological opportunities and challenges
This section focusses on the technological opportunities and wider challenges involved in developing virtual reality tools for chemical engineering education. An overview of the generic J o u r n a l P r e -p r o o f 14 virtual reality-based interface that can be developed and deployed for chemical engineering education is shown in Fig. 2. Firstly, any computer-aided design (CAD) software is used to prepare three-dimensional models of different process equipment and operating environment. Further, commercial virtual reality software is used to create immersive three-dimensional virtual environments, for which the CAD model is given as an input. The virtually immersive experience can be enhanced with the help of accessories including virtual reality headsets and optical sensors.
Further, the virtual reality interface can be augmented with mathematical models to develop interactive learning activities for the purpose of education as evident from Fig. 2.

Technological opportunities
As stated earlier, virtual reality is mainly used in chemical engineering education as a visualisation environment for viewing 3D models of process plants. Moving beyond visualisation, such tools can be augmented with mathematical models to enhance the immersive learning experience.
In the domain of chemical engineering education, specific applications of mathematical models include the virtual operation of dynamic process plants, the testing of hazardous scenarios, and, the simulation of potential explosions in a chemical reactor . In this regard, either first-principle models, data-driven models, or a set of hybrid models can be used to develop dynamic process simulations and interactive learning modules. An overview of the different modelling approaches and their key features for virtual reality-based systems is depicted in Fig 3.

Data-driven models
In recent years, there has been an increasing trend towards using data-driven models (also known as machine-learning models) in process engineering. One example of a data-driven model is a digital twin model. A digital twin is a virtual model of a physical process or a system, that can be used to monitor the physical entity in real-time . Digital twins are widely used in industry as a process-monitoring tool for predicting faults and accidents, and for real-time optimisation of processes .
Such models can also be coupled with virtual reality systems to develop reality-simulated learning and training modules. The combination of virtual reality and digital twin technologies enables realtime data exchange between the actual process and the virtual replica. Therefore, such a co-J o u r n a l P r e -p r o o f 16 simulation environment captures the behaviour of processes and can be utilised for predictive modelling studies .

Hybrid models
Moving beyond the conventional first-principle models and data-driven models, it becomes necessary to develop hybrid models so as to effectively capture the complexity and uncertainty of real-world processes . Such hybrid approach often makes use of a combination of first-principle models and data-driven models to enhance the modelling capabilities of immersive process plants.
The combination of models can be used to develop various realistic learning scenarios for testing by using virtual reality-based educational systems.
It is to be noted that the output from mathematical models are often quite abstract in nature. In this regard, virtual reality can be augmented with mathematical models to enhance the immersive learning experience. It is worth noting that first-principle models can be used to simulate the real behaviour of process plants. On the other hand, simulating virtually unsafe scenarios and the prediction of process equipment faults can be better captured whilst using data-based models.

Technological challenges
Virtual reality technology is at an early stage of adoption in the domain of chemical engineering education. Some of the associated technological challenges are discussed below: (3) Technical functionality: Education using virtual reality technology requires advanced hardware and computer systems. Therefore, the unavailability or failure of such devices may hamper the intended learning experience designed by the teachers .
This section has emphasised the need for augmenting virtual reality tools with mathematical models for advanced educational applications, which would allow the users to view the model and the modelling output in a virtually immersive environment. It also highlighted few technological challenges, primarily related to system construction and maintenance, that may hinder the ability of an educational institution to deliver virtual reality learning solutions on an ongoing basis.

Educational implications
The wider pedagogical implications of virtual reality for chemical engineering education are discussed here. Virtual environments are being widely employed in educational settings, with studies showing potential teaching and learning benefits resulting from the utilisation of such environments . Furthermore, interactive virtual reality-based environments can offer greater benefits due to the active rather than passive mode of learning .
In the domain of chemical engineering education, the use of virtual reality is expected to have the following noteworthy benefits: (1) Students or operators can access the virtual interface without any space and time related constraints. In addition, it allows the user to control the movement of the virtual camera J o u r n a l P r e -p r o o f 18 within the interactive learning environment . This permits the users to learn at their own pace unlike in a real process plant.
(2) In the case of a virtual process equipment (see Section 5), it is possible to 'cut' open and observe the internal operations of the equipment. This might not be practically feasible while using a real physical equipment. A real equipment is often insulated and thus, it can be viewed only as a 'black box' piece of equipment, as evident from Fig. 4.
(3) Virtual reality allows the users to learn in a safe, controlled and risk-free environment. This includes replaying and experiencing various unsafe and hazardous scenarios. This might not be practically feasible by using a real chemical plant.

(4) Process operators can experience various what-if scenarios that requires teamwork and
coordination . This can be achieved by creating virtual avatars or digital representations of operators to experience teamwork during virtual plant shutdown, emergency, and evacuation scenarios . Similar multiplayer activities also make the undergraduate students more equipped to take up industrial positions after their graduation.
(5) Development of the learning content can be better, faster and continuously updated by using active student involvement not only as users, but also in planning and creating the necessary content. This allows the teaching activities to reach significantly higher levels of Bloom's taxonomy . The skills acquired from learning at higher levels of Bloom's taxonomy are increasingly important in the present context, where information is easily accessible, while active planning and creative utilisation of that information are even more desirable skills, not least in a chemical engineer. For example, developing a library of chemical process design attempts in the virtual environment, including successful and failed attempts, would be beneficial to both the student and the teacher.
J o u r n a l P r e -p r o o f 19 (6) A VR environment provides excellent opportunities to make use of peer instruction , as students' VR creations (e.g. a chemical process design) can be made completely anonymous and evaluated by the student peers. This reduces the teacher load while increasing the student learning, e.g. as assessed by higher levels on Bloom's taxonomy.
Peer instruction has also been demonstrated to minimise differences between students with different educational backgrounds , which enables an efficient mode of teaching particularly at the postgraduate level, where the student base is typically quite varied and international.
At the same time, there are certain drawbacks associated with virtual reality-based education, as discussed below: (1) VR might not be an appealing educational platform for all users. For instance, it might not be feasible for those users who are unable to appreciate any type of learning that involves interactive educational activities . For such users, immersive VR might only be able to create situational interest on a topic and may not encourage deeper learning ." (2) Another drawback is that the use of virtual reality devices such as HMD could cause physical discomfort including cyber sickness, nausea, headache, and eye strain to the users . This might also distract the users and could negatively impact the immersive pedagogical experience.
The review of the existing case studies in the literature has exposed the following pedagogical  The educational impact of VR learning tools has only been partially explored so far. It is worth mentioning that traditional learning design has been used for decades and indeed centuries, while VR educational design is still in its infancy. In addition to enabling training in situations that would otherwise be harmful and/or entail material wastage, VR educational design enables interactive visual learning experiences, which are almost impossible with traditional learning design for various reasons. For example, in virtual environments students can have experiences that are physically impossible such as visualising and manipulating very small objects (e.g. molecules or cells) or very large objects scaled up or scaled down (e.g. a full scale chemical process plant) so that it becomes easier to explore and to understand. In addition, such environments allow the students to have unlimited repetition of procedures and tasks to be learned, such as standard operating procedures (SOPs) or safety procedures, which is not possible in traditional learning design due to practical constraints. Finally, virtual and digital learning environments can cater to a wide range of student users unlike traditional learning environments. For example, while traditional learning environments in engineering primarily accommodate engineering students, virtual reality learning can also accommodate younger students, lifelong learners, as well as students that cannot physically access, or cannot afford, traditional learning environments. The societal impact of this democratising feature of the virtual learning experience is yet to be assessed.
Therefore, further research is needed to evaluate the wider educational impact of virtual reality on the student user community.

Economic and social challenges
The inclusion of virtual reality for education is expected to provide additional learning and training benefits, as discussed in the previous section. However, the additional benefits come along with J o u r n a l P r e -p r o o f 22 additional costs for developing the virtual environment . Since virtual reality is an evolving mode of education, it becomes necessary to discuss the wider social and economic implications on the user community.
This section touches upon the social and economic implications of virtual reality for education with the help of an ongoing case study application. In work progressing at the Eindhoven University of Technology in the Netherlands, a virtual reality model of an existing fluidised bed membrane reactor is being developed as an executable program that can be run on a personal computer. The key objective of this case study is to develop a virtual reality platform that can be used for teaching and training of undergraduate chemical engineering students at the university.
Further, virtual reality headsets along with hand controllers are used to experience the immersive learning environment. An overview of the existing real membrane reactor at the university laboratory and its corresponding virtual reality twin is shown in Fig. 4.
The development of the virtual membrane reactor requires advanced hardware as well as gaming software. The associated capital expenses can be split into three components as follows: (a) Cost for virtual reality hardware (b) Cost for virtual reality software (c) Human resource costs The hardware for this study primarily consists of the desktop and virtual reality headset. The cost breakdown for the different hardware components is shown in table 3. The interface is being developed using the Unity 3D game engine with the backend programming done in C# language.
Further, the SteamVR Unity Plugin is used to setup the hardware and software interface . . The personal free version of Unity that is available for educational and non-business purposes is used for the interface development. It is to be noted that the license terms for the personal free version The long-term uninterrupted utilisation of virtual reality tools would also require operating expenses in the form of technical maintenance and software support. However, these expenses are difficult to quantify and are not presented in this paper.
This case study application assumes that there are 20 students in a classroom, divided into groups of 4 people. This would require five sets of VR hardware and five gaming desktops as shown in Table 3. The availability of multiple VR systems allow the student groups to experience and interact with the immersive virtual environment at the same time. The resulting overall expenses and the costs involved per student are presented in Table 5.
The virtual exercises that are being developed include a self-guided tour of the fluidised bed membrane reactor for hydrogen production, a piping and instrumentation diagram (P&ID) referencing interface, a virtually unsafe reactor scenario, and a technical quiz for learning evaluation. Further, teachers could gather the students' opinion about the developed virtual interface with the help of a survey. The different questions proposed as part of the survey are listed in Table 6.
J o u r n a l P r e -p r o o f 24 It is worth noting that this is a generic case study application and the estimated costs would change on a case to case basis. Nevertheless, it provides an overview of the additional expenses involved in developing an immersive VR environment for chemical engineering education.
From an economic perspective, the availability of financial resources continues to be a major barrier for the massive adoption of VR for education. Most of the educational institutions have limited funding and are mostly reluctant to invest in expensive VR gadgets when compared to traditional classroom or laboratory-based learning . Moreover, the cost-effectiveness of using VR for education is difficult to quantify and needs to be evaluated in terms of additional costs to additional benefits. However, the benefits or the returns of investment for education are yet to be quantified . One possible benefit is that learning with a virtual reactor, for example, does not require the use of actual chemicals. Therefore, such education helps to avoid undesirable wastage of chemicals. In addition, the users can interact with the virtual model multiple times, making it ideal for the purpose of operator training.
There is a lot of cross over in terms of the social and pedagogical benefits of VR-based education interfaces. The ready accessibility and availability of VR teaching interfaces is one such example.
At the same time, there are certain social challenges associated with the slow adoption of VR tools for education.

The readiness of teachers and students to embrace new technologies for education is a key
challenge. An assessment of user readiness for technology-based classroom learning is necessary before the implementation of immersive virtual environments for education. This is also important to ensure long-term motivation for using virtual reality among the learning community. In the near future, the virtual membrane reactor, referred to above, will be tested with a view to qualifying its value for undergraduate chemical engineering education. During testing, it will be compared with laboratory experiments performed in a real reactor, which can be documented using the traditional laboratory notebook concept. The level of learning will be assessed by conducting a technical quiz on the membrane reactor before as well as after performing the virtual exercises. Further, a methodology will be developed to evaluate the educational impact of virtual reality on the undergraduate student community.

Conclusion and future directions
The paper reviewed the state of the art in virtual reality technology for chemical engineering education. The review highlights that there is an increasing scholarly interest towards utilising virtual reality in chemical engineering academia and in industry. Various studies have made efforts J o u r n a l P r e -p r o o f 26 to develop virtual process plants that can cater to the educational needs of both industry and academia. The paper identified opportunities and shortcomings in the fundamental areas of technology, pedagogy and socio-economics for virtual reality-based education. The paper emphasised the need for augmenting virtual reality interfaces with mathematical models and discussed the modelling approaches present behind such systems. Further, the paper identified the major pedagogical strengths and limitations associated with such education. The paper also stressed upon the need for novel educational impact assessment methodologies to evaluate virtualreality-based learning. Further, a case study was presented to discuss the social and economic implications, and to identify the bottlenecks associated with the slow adoption of virtual reality tools across chemical engineering education.
The review has the following limitations due to the reduced scope. Firstly, the review focussed on virtual reality-based case studies rather than just virtual laboratories. This was done due to the overall research motivation for this paper. Another weakness is related to the quantification of learning outcomes for the different case studies. Almost all the studies evaluated the learning only in terms of user feedbacks on different virtual reality accessories and the virtual environment.
Finally, all the studies had a strong bias towards the positive aspects of virtual reality in chemical engineering education. Therefore, it was difficult to discuss the drawbacks of existing case studies.
Nevertheless, the above limitations do not diminish the value of this paper, and in fact, highlight the opportunities for future research in this field. As part of the future work, the following recommendations are made.
Firstly, a quantitative cost versus benefit analysis could be carried out. In this regard, the training benefits can be quantified in terms of the avoided financial costs resulting from plant failures and shutdowns owing to poor training of the users. The capital expenses and the monetary benefits can J o u r n a l P r e -p r o o f 27 be used to determine a cost to benefit ratio for the virtual process and for the real process. This ratio would aid in decision-making between the two modes of education from a socio-economic perspective.
Secondly, virtual reality-based education has certain strengths and limitations as discussed in section 4. For example, it cannot be used to directly display and visualise mathematical formulations. In such situations, traditional classroom-based approach becomes more effective.
On the other hand, traditional learning might not be engaging and inspiring enough for students to attend classroom lectures. Therefore, future case studies should explore options that utilise a combination of traditional classroom-based learning and virtual reality-based learning. This kind of a hybrid approach would help to reap benefits from both modes of education.
Finally, it can be concluded that virtual technologies have become more promising during the COVID-19 pandemic. In this regard, virtual reality can be regarded as a powerful tool for remote learning in chemical engineering education. For instance, Seifan and co-workers developed a virtual field trip for the undergraduate students of engineering biotechnology course at the University of Waikato. It is acknowledged that such experiences cannot replace physical industrial field trips. However, such virtual technologies are the best possible alternatives during situations where the real experience is not practically feasible.

Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.     Table 6: List of questions proposed in the feedback survey Question Opinion (Strongly agree = 4, Agree = 3, Neutral = 2, Disagree = 1, Strongly disagree = 0) Could you clearly visualise and understand the different parts of the membrane reactor that was taught earlier in the classroom? Do you believe that the VR interface improved your understanding of the membrane reactor when compared to the real reactor in the laboratory? Do you feel that audio narration would improve the learning during the self-guided tour?
Did you experience any sort of physical discomforts such as nausea, headache, or eye strain while using the VR hardware? Do you think that VR can be used as an alternative mode of education at the Eindhoven University during the COVID-19 pandemic? Do you believe that a hybrid model involving traditional classroom learning, hands-on experiments, and VR exercises would be the best possible approach for education?