**Digital Twin for Variation Management: A General Framework and Identification of Industrial Challenges Related to the Implementation**

#### **Kristina Wärmefjord 1,\* , Rikard Söderberg <sup>1</sup> , Benjamin Schleich <sup>2</sup> and Hua Wang <sup>3</sup>**


Received: 20 April 2020; Accepted: 8 May 2020; Published: 12 May 2020

**Abstract:** Digital twins have gained a lot of interest in recent years. This paper presents a survey among researchers and engineers with expertise in variation management confirming the interest of digital twins in this area. The survey shows, however, a gap between future research interest in academia and industry, identifying a larger need in industry. This indicates that there are some barriers in the industry to overcome before the benefits of a digital twin for variation management and geometry assurance can be fully capitalized on in an industrial context. To identify those barriers and challenges, an extensive interview study with engineers from eight different companies in the manufacturing sectors was accomplished. The analysis identifies industrial challenges in the areas of system-level, simulation working process, management issues, and education. One of the main challenges is to keep the 3D models fully updated, including keeping track of changes during the product development process and also feedback changes during full production to the development engineers. This is a part of what is called the digital thread, which is also addressed in this paper.

**Keywords:** digital twin; manufacturing; tolerancing; geometry assurance; digital thread

#### **1. Introduction**

Variation is an unavoidable element in mass-production. If not handled properly, variation can cause problems and significantly increase costs. Variation management is a broad term, relating to different methods to handle, and reduce the effects of, variation. In this paper, the focus is on geometrical variation. Problems related to geometrical variation usually constitute a significant part of the total cost for poor quality, sometimes up to 40% of the total cost for a manufacturing company in the form of delays, scrap, repair, rework, unsatisfied customers, and warranties [1,2].

Methods to reduce the effects of geometrical variation are sometimes referred to as geometry assurance. Additionally, terms such as tolerance analysis and tolerance management are used. Geometry assurance activities can be executed in different phases of the product realization cycle and cover areas such as the design of locating schemes/fixture layout, variation simulation, tolerance analysis, and inspection.

The digital revolution is expected to have a huge impact on the manufacturing industry in the coming years. Industry 4.0, increased digitalization, and access to an increased amount of data also open up for new methods and tools for geometry assurance, such as real-time process optimization using digital twins [3].

This paper aims at, from a standpoint in the current and future research areas and industrial needs, to identify the challenges, mainly in terms of processes and data flows, that need to be in place to fully capitalize on the possibilities offered by digital twins for zero-defect manufacturing.

A survey was distributed to experts in variation and tolerance management in academia and industry to identify the gap between research areas and needs, today and in the future. Not very surprisingly, this survey pointed out an increased demand for digital twins for improved geometrical quality. A deeper analysis using on-site interviews with over 40 geometry engineers from eight Swedish and Danish companies within the manufacturing sector was then conducted. The purpose of this more detailed study was to increase the understanding of the industrial needs and challenges related to the implementation of digital twins. Those results are the main contribution of the paper.

In the remaining part of the introduction, the basics of geometry assurance are described.

#### *1.1. Geometry Assurance*

Geometry assurance is a term describing a set of activities, all aiming to improve the geometrical quality of an assembled product. The geometrical quality affects both the esthetical and functional requirements of the final product. Usually, the maximum geometrical variation allowed in a certain dimension on the product is specified with a tolerance.

The geometry assurance activities start in the early design phases of the product realization process with tasks such as locating scheme optimization to secure a robust positioning system during the joining of parts and variation simulations to predict the variation in the critical dimensions of the final assembly. Inputs to the variation simulation are part geometries, information about joining and fixturing and inspection data, or suggested tolerances on the part level. The analysis is done iteratively to find a reasonable set of part tolerances, i.e., finding limits for maximum allowed variations on the part level at an acceptable cost. The simulation is sometimes based on the assumption that parts are rigid [4], and sometimes, the flexibility of, for example, sheet metal parts is included [5,6]. The latter one usually improves the simulation accuracy [5] and makes it possible to include clamping and joining forces, different material characteristics, joining processes, etc.

The variation simulation can be done in a standalone CAT (computer-aided tolerancing) software, such as RD&T [7], or integrated into the CAD system, as, for example, 3DCS [8]. Both setups are widely used industrially and have their pros and cons.

#### *1.2. Scope of the Paper*

This paper focuses on pinpointing future research and industrial needs within the area of variation management and geometry assurance. Digital twins are identified as an important area, and challenges, limitations, and requirements related to the implementation of digital twins for geometry assurance in the industry are identified and discussed.

In Section 2, the survey aiming to identify gaps between current and future research areas and needs is presented. Digital twins for geometry assurance are discussed in Section 3, and the interview setup and methods are described in Section 4. The findings are presented in Section 5, and challenges related to the implementation of a digital twin in an industrial context are highlighted in Section 6. This is followed by discussions and conclusions in Sections 7 and 8, respectively.

#### **2. Materials and Methods**

As a starting point for this work, a questionnaire was distributed among experienced researchers and engineers in the areas of tolerance analysis and geometry assurance. The purpose was to identify what research topics are perceived as most important today and in the future and, also, what the industrial status and needs are today and in the future.

## *2.1. Questions*

A number of questions considering background variables were posed. Those questions gathered information about age, gender, years of experience in tolerance analysis, affiliation (industry or academia), and what continent the respondent lives on. This information, together with the answers to the four questions below, will be used for the analysis in this section:


The keywords to choose between were the same for all four questions and can be seen below.


#### *2.2. Statistical Analysis and Method*

This digital questionnaire was distributed among participants of the European Group of Research in Tolerancing (E-GRT) Biannual Seminar 2019 and other researchers and industrial engineers with an expertise in tolerance analysis and variation management. In total, 43 answers were collected. Information about the background variable responses can be seen in Table 1. For age and years of experience in tolerancing, the respondents could choose between different alternatives presented in the form of intervals. The midpoints of the intervals were used in the calculation of the average values presented in Table 1.



The answers to the questions posed above were analyzed with the purpose to see the participants' opinions about current focus areas and future needs in research and the industry. To analyze the frequency differences among different answers, a Cochran Q test was employed. This statistical test is a nonparametric test for related categories with binary responses [9]. The related categories are the questions (a)–(r). This test can determine if there is a significant difference between the categories. The test statistic, for binary responses *Yij*, *i* = 1, . . . , *n*, and *j* = 1, . . . , *k*, is:

$$Q = \frac{(k-1)\left(k\mathbb{C} - T^2\right)}{kT - R} \tag{1}$$

The variables are: *n*: number of respondents, and *k*: the number of questions.

$$\mathcal{C} = \sum\_{j=1}^{k} \left( \sum\_{i=1}^{n} Y\_{i,j} \right)^2$$

$$T = \sum\_{i=1}^{n} \sum\_{j=1}^{k} Y\_{i,j}$$

$$R = \sum\_{i=1}^{n} \left( \sum\_{j=1}^{k} Y\_{i,j} \right)^2$$

The *Q* statistic follows approximately a χ 2 -distribution with *k* − 1 degrees of freedom. This approximation is considered valid if *n* > 4 and/or *nk* > 24. Here, *n* = 43, and *k* = 18.

The Cochran *Q* test is used to test whether there is a difference between any of the different categories. To point out what categories significantly differ from the other, a minimum required difference (MRD) test is used with Bonferroni adjustments to compensate for repeated tests [9]. The MRD is calculated as:

$$MRD = z\_{adj} \sqrt{2 \frac{kT - R}{n^2 k (k - 1)}} \tag{2}$$

where *zadj* is the value of (1 − α*adj*/2) in a standard normal distribution. The MRD is based on pairwise observations.

When comparing answers to the same question, with filtering based on the affiliation of the respondent, another test needs to be used. In this case, a standard *z*-test with pooled variance will be used [10]. This test compares the proportions, *p<sup>i</sup>* and *p<sup>j</sup>* , respectively, of positive answers to a certain question and test the hypothesis

$$H\_0: p\_i = p\_j$$

for two groups, *i* and *j*.

#### *2.3. Results*

Using the Cochran Q test previously described, it can be shown that there are significant differences between the alternatives (a)–(q) for all four questions: (1)–(4). The *p*-values for each question are all below 10−<sup>7</sup> . For a deeper analysis of what categories (a)–(q) actually have a significant difference, the MRD is applied.

In Figure 1, the results from the survey are shown. Some observations can be made.


To see the differences between the situation today and future needs, the differences between the answers are compared (see Figure 2). This illustrates the increased need of (f) (digital twins), g − /2)

(zero-defect manufacturing), k (new manufacturing methods), and n (Industry 4.0), especially in the industry. For digital twins (f) it can be noted that the increased need for industrial development is not matched by an equally large increase in future research. Digital twins are already a hot research topic, as seen in the left part of Figure 1, and the challenge is the applicability of the digital twin framework in the industry. This will be the focus of a more in-depth interview study, presented in the following sections. – – – twins (f). For simulation (e), the MRD test shows that there is a significant difference (α 

: =

<sup>0</sup>

–

 = ∑ (∑ , 

=1

 = ∑ (∑ , 

=1 2

= √2

=1

=1

=1

=1

 = ∑ ∑ , 

2

)

) 2

−

<sup>2</sup>( − 1)

−

**Figure 1.** Responses to questions (1)–(4). The left part shows what areas are believed important today in research (blue) and the industry (red). The right part shows what areas are believed to be important in the future.

future's and today's research and development needs in research (blue **Figure 2.** Gap between the future's and today's research and development needs in research (blue bars) and the industry (red bars).

Other interesting observations from the survey can be reached if the data is filtered based on the affiliation of the respondents (see Figure 3).

*α* **Figure 3.** Responses filtered based on the affiliations of the respondents. Significant differences, using α = 0.10, are encircled.

*α* By filtering data into groups, the sample size for each group decreases, and therefore, not all differences are statistically significant. The z-test introduced in Section 2.2 is used to test for significance. Significant differences, using α = 0.10, are encircled in Figure 3.

Some observations about the future:


In general, it can, based on the study, be concluded that digital twins are one of the most important areas within variation management. Today, it is most important in research, and in the future, it will become very important also in the industry, with an increased industrial need for development of the digital twins. In the next section, the meaning of a digital twin from a geometry assurance point of view is explained.

#### **3. Digital Twin Framework for Geometry Assurance**

The digitization of manufacturing is increasing. In the manufacturing business, 68% of the companies in a recent study stated that digital manufacturing is a top priority [11]. The number was even higher in countries such as India (94%) and China (87%). It is, however, also reported that the expected benefits are perhaps not yet fully reached. In the context of Industry 4.0, digitals twins are seen as one of the top ten technology strategic trends, according to Gartner Research [12], and various potential applications of digital twins along the product life-cycle have been identified [13,14].

A digital twin is a digital replica of a physical entity. In manufacturing, the terms digital model, digital shadow, and digital twin are used to indicate different levels of data exchange [15]. A digital model is a digital representation of a physical part or assembly without data exchange, a digital shadow

enables a one-way data flow between the physical and virtual representations of an object, while a digital twin allows for a bi-directional automatic data exchange between the physical and virtual representation. In this regard, different reference models for digital twins have been proposed, such as in [16], considering important properties, such as model scalability, interoperability, expansibility, and fidelity. Most likely, a family of digital twins is needed to achieve optimized products and production flow, where each twin has its own purpose. The digital twins must represent both the product and the production system. As stated in [17], multi-scale analyses and distributed decisions will be required for optimizing the activities on the shop floor, as well as processes, resources, and machines. In many cases, digital twins will be interlinked. For example, the digital twin for geometry assurance described below must be linked to twins controlling the assembly cell equipment.

For the geometry assurance application, a digital twin approach was presented by Söderberg et al. [3]. This approach is also adopted in this paper. It was suggested that the simulation models, used in the design phases, can be reused and fed with inspection data during full production to do real-time individualized optimization of the assembly process to reach high-quality products. In Figure 4, an overview of the idea can be seen. –

**Figure 4.** A digital twin for geometry assurance [3].

The idea is that two parts, A and B, are inspected using, for example, 3D scanning. This data is fed into the simulation model, which constitutes the kernel of the digital twin, to match individual parts over a batch of parts to minimize the geometrical deviation and variation of the assembly of A/B. This is also referred to as selective assembly [18]. Furthermore, the joining process can be optimized with respect to geometrical quality. Locators can be adjusted [19], and spot welding sequences can be optimized [20] for each individual assembly. By those suggested adjustments, the geometrical quality of the assembly can be improved without changing the tolerances of the parts, which is beneficial from a cost perspective.

*—* With this approach, studies on several industrial cases have shown that the variation of the final subassembly can be reduced up to 50%, compared to when no adjustments are done [18–20].

With the labeling digital model, digital shadow, and digital twin mentioned at the beginning of this section, a digital twin for geometry assurance can be interpreted as:


Those three pillars constitute together a digital twin.

#### *3.1. Digital Twin—The Digital Model*

As stated in [21], modeling and simulation are key aspects to implement a digital twin. To build the digital model, which is the core of the digital twin depicted in Figure 4, information about the joining and assembly process, part geometries, and other characteristics are needed. Using this information, a variation simulation model can be built. In the approach suggested in [3], the commercial software RD&T was used. This digital model is capable of relating deviation and/or variation on the part level to the deviation and/or variation on the assembly level.

Nominal dimensions, surface textures, material data, etc., which can constitute the input to build the digital model and complement the 3D model of a part, are sometimes referred to as product manufacturing information (PMI), according to ISO 16792. The use and development of PMI are, to some extent, driven by the fact that the 3D models successively are replacing 2D drawings [22]. This is beneficial not only due to its potential to reduce the amount of time dedicated to producing 2D drawings but also since the accuracy is supposed to increase.

Hedberg et al. [23] and other researchers at NIST (National Institute of Standards and Technology) discussed a digital model-based definition (MBD) for engineering tasks in manufacturing and inspection phases. This relates to the concept of a digital thread, which referrers to the digital information and data flow between different product realization phases. The digital thread provides important information to the digital model.

The 2017 ISO 1101 tolerancing standard allows for more precise tolerance definitions, which are independent of the viewing plane, and also supports the digital thread.

From a geometry assurance perspective, the digital thread must allow for a cohesive digital information flow between all activities related to product geometries, requirements, fixtures, and assembly cell layouts. This will be the focus of the interviews presented in Section 4.

#### *3.2. Digital Twin—The Input Data*

Scan data of the individual parts constitute the input to the digital model and can also be seen as a part of the digital thread. Given part deviations, adjustments adapted to a certain set of parts, A/B, can be predicted using the digital model. Of course, the scan data needs to be accurate and reliable. Scan data can be mapped to the nominal finite element meshes of the parts. Aspects of the scan data as the input to a digital twin for geometry assurance are discussed in [24] and are not the main focus of this paper.

#### *3.3. Digital Twin—The Output Data*

By feeding the digital model with inspection data, optimal locator adjustments and optimal spot-welding sequences can be determined. This constitutes the output from the digital model and is, of course, an important part of the digital twin. A suggestion of how to handle the data flow between the digital model and the assembly cell is presented in [25].

#### **4. Interviews**

To clarify the industrial state of the art and future needs regarding geometry assurance and digital twins, semi-structured interviews with over 40 engineers at eight Swedish and Danish companies in the manufacturing sector have been conducted.

The presented findings are focused on the information needed to build the digital model, using the terminology introduced in the previous section. However, to not limit the scope of the interviews and risk missing important aspects of the digital model and the simulation procedure, the whole geometry assurance loop is addressed. The importance of acknowledging interactions between different systems in a digital twin context is also highlighted in [26].

The engineers participating in the interviews work with geometry assurance or on a more general level with product development. Managers responsible for geometry assurance at the different companies were contacted, and they helped to select suitable interviewees. The interviewees were chosen based on their high competence and long experience in the area.

The base assumption of the interviews is that there exists a product development process containing concept, planning, and production phases. At most companies, this process is further detailed with gates and subphases [27,28].

A digital model, which will be the core of a digital twin, should be developed, reused, and updated in the different phases of the product development process in order to provide as much value as possible. Therefore, the respondents of the interviews were asked about the current situation regarding different gates and goals related to geometry assurance in the different phases (see Figure 5). They were also asked about the responsibilities of different roles, about input and output data for the different activities, and what (software) tools they use. The guide in Figure 5 was filled out together with an identical form reflecting what an ideal situation for the geometry assurance activities should look like in the future.


#### **Figure 5.** The interview guide.

As a part of the preparations for the interviews, the form, used as a guide during the interviews, was developed according to the recommended steps for a semi-structured interview described in [29]. Those steps are:

(1) identifying the prerequisites for the interviews.

In this step, the general procedure for the interviews was outlined. The ability to focus on issues being meaningful for the participants and also to allow diverse perceptions to be expressed were taken into consideration [30]. The companies participating in the study were contacted. All contacted companies accepted the interview invitation.

(2) retrieving and using previous knowledge.

Previous research in the area was investigated. The two interviewers had long (over 15 years) experience of research in the area.

(3) formulating the preliminary semi-structured interview guide.

The form illustrated in Figure 5 was developed. This form provided a framework for the interview and allowed easy movement from question to question, which was beneficial for the results [31] but, also, the possibility to dive deep into certain topics.

(4) pilot testing the interview guide.

To confirm the relevance of the developed interview guide, the questions were tested on an experienced engineer with many years of experience as a consultant within geometry assurance. Some small adjustments were made to make the questions clearer.

The interviews were held by two interviewers to reduced bias introduced during the interview. Before the interview started, its purpose and format were clarified. The respondents were free to elaborate on the questions, and follow-up questions were used to clarify answers.

After the interviews, the results were analyzed. One person took notes during the interviews and immediately afterward, those notes were clarified and validated with the other interviewer.

#### **5. Results: The Geometry Assurance Process Today**

In this section, the results from the interviews reflecting the present situation for the three main product realization phases are presented. An overview of the results can be found in Figure 6. In the figure, the different activities (middle) are listed together with the input required for the activity (top), as well as the output generated from the activity (bottom).

**Figure 6.** The input and output to the current digital thread for different geometry assurance activities.

#### *5.1. The Concept Phase*

For most companies in the study, the geometry assurance process starts with some kind of evaluation of concepts suggested by the design department. The suggested concepts are based on customer demands and other requirements and, to a large extent, also on experiences from previously produced models and projects.

① The geometry engineers check if the suggested concept(s) can be built with reasonable geometrical quality (see O<sup>1</sup> , Figure 6). At this first stage, the evaluation is usually based on discussions, previous experiences, and sometimes, also on simple styling surfaces. The positions of the split-lines between parts affect both the visual sensitivity to variation and the level of geometrical variation in, for example, a weld between two parts [32].

② The choice of locating scheme (O<sup>2</sup> , Figure 6) is critical for the geometry assurance process. The locating scheme, used to fixate parts during assembly, must lock all degrees of freedom of a part and should be as robust as possible to variation. At this stage in the process, there usually exists a (not-so-detailed) CAD model of the part. This is exported from the CAD system in a neutral triangular format (often WRL or JT) and used in a variation simulation software to evaluate robustness. At the companies in this study, RD&T [7] is the most commonly used software. The suggested locating schemes are documented in reports and 2D drawings generated from the CAT tool. At one of the companies in the study, the produced results were linked to the 3D digital model.

③ The next main step is to define the tolerances on the parts (O<sup>3</sup> , Figure 6). The goal is to find tolerances such that the final requirements on the product can be fulfilled but, at the same time, keep costs down. Again, triangular representations of the CAD model are used as the input, together with

old inspection data (from similar concepts), if available. The analysis is done using variation simulation software to predict the final variation on the product (or subassembly) level or by simplified analyses in Excel or similar. The predicted variation is compared with the requirement on the product level, and the process is iterated until satisfactory results are obtained. Of course, experiences from previous projects are also important input. The output is the part tolerances described in a variation simulation model and/or on 2D drawings.

#### *5.2. The Planning Phase*

In the planning, or verification, phase, the interviews focused on system verification activities from a geometry assurance perspective and on inspection preparation.

The system verification (O<sup>4</sup> , Figure 6) is often done through physical tests and try-outs. This is to complement and extend learnings from simulations. Of course, it would be desirable to reduce, or even exclude, the physical tests and rely only on simulations. This is a matter of "virtual trust" (i.e., trust in simulated results), a topic discussed in the following sections. Changes are done to increase producibility and quality. The changes can include minor changes of the geometries and changes of tolerances, fixtures/locating schemes, joining sequences, etc. After those changes, it is of course very important to update the CAD models and other information in the digital model and the digital thread.

One of the companies reported ongoing work to replace physical trimming, where a part is positioned in a checking fixture to control the shape of the flanges, with a virtual method based on the variation simulation model [33]. This is something that most companies want to do, but most of them are not there yet. Physical tests are expensive and should be kept to a minimum. At one of the bigger companies, it was reported that over 180 persons work with handling the most acute geometry-related problems exposed during system verification.

The inspection preparation (O<sup>5</sup> , Figure 6) is an activity where the practices at the companies differ a lot. Some companies set inspection points according to the final requirements defined earlier and complement this with process points to monitor the process. At other companies, the chain from the final requirements to inspection points is not that clear. The inspection point preparation is done by the design engineer. Other aspects mentioned were that it is bad practice to do an inspection point preparation leading to points that cannot be evaluated/measured in a good way, that it is difficult to measure very flexible parts (due to effects from gravity), and that the number of inspection points often is too large and lead to difficulties in the evaluation and storage of data. Inspections are also associated with a cost, and it is desirable to keep the number of inspection points to a minimum. At the more advanced companies, 3D scanning is mixed with a CMM (Coordinate Measuring Machine) inspection. At other companies, gauges and other manual inspection tools are used.

For the digital thread, it is important to update the 3D models with inspection point coordinates. This is not always done. Quite often, an inspection point report is generated in PDF or similar, but the information is not fed into the CAD system.

#### *5.3. The Full Production Phase*

During full production, the geometry assurance activities are supposed to subside. Most of the geometry assurances should be preventive work in the early phases to make sure that the product and the processes are robust to variation. There are, however, always issues arising, and root causes of deviations and variations must be identified, and changes in the process should be documented (O<sup>6</sup> , Figure 6).

Often, design changes done during full production are not brought back to the design engineers in a systematic way. There can be local systems, logs, and documentation, but it is not fully integrated into the design process. Therefore, it is a risk of repeating mistakes in future projects. Moreover, there is a disagreement between the virtual models and the physical world, and the generated inspection data will not fully reflect the virtual models. This has also been addressed in [34].

The importance of a good inspection database was mentioned at several companies. They used systems like QSYS [35] and CM4D [36] but were still missing some company-specific features. It can be concluded that the inspection database needs to be adapted to specific company requirements but, also, that different users have different needs and demands.

## **6. Results: Challenges Related to the Implementation of a Digital Twin**

The future needs, identified during the interviews, that should be addressed to achieve a comprehensive digital twin for geometry assurance, are listed below. They can be divided into four different categories: system level, simulation working process, management issues, and education. The needs are further discussed in Section 7.

## System-Level

(a) The major future needing identified at all the companies is a fully updated 3D digital model developing through all the phases in the product realization process. This is outlined in Figure 7. All activities and changes that affect the geometry of a part should also be brought back to the 3D digital model of the product, securing an always-updated and true geometrical representation of the actual produced part. This is today not the case at the companies in the study.

The 3D model should be updated when new information is available during the product realization process. Besides changes in geometry, locator positions, tolerances, and inspection point positions should also be linked to and updated in the 3D model.

Simulation Working Process


#### Management Issues


## Education

(k) Education at universities should cover tolerancing and knowledge in geometry assurance to a greater extent. This finding is also supported by unstructured interviews with experts from German automotive suppliers, who reported that more attention should be paid to education in tolerancing and geometry assurance at universities. More particularly, the students should be introduced to the different steps of the geometry assurance process (as highlighted in previous sections) to fully understand the importance, interdependencies, and repercussions of tolerancing decisions on product quality and cost. Additionally, to cope with the challenges of a digitized design and manufacturing environment, students will need more knowledge and competencies regarding model-based definition workflows and the digital thread in geometry assurance in the future.

**Figure 7.** The CAD model needs to be updated with all new information from different geometry assurance activities (and other activities) to reflect the as-fabricated status.

#### **7. Discussion**

— — Among the identified needs listed in Section 6, the main need related to increase the use of digital twins and the digital thread supporting them is point (a) on the system level, the continuous update of the 3D models. If the 3D models only reflect the as-planned and not the as-fabricated product data, there is a huge risk for costly mistakes when using them in a digital twin concept. It is also difficult to learn from previous projects and improve new products without an updated 3D digital model. The reason for this lack of updated models might be practical, since some geometry assurance (and other) analyses are based on neutral triangular formats, such as WRL or JT, and the output might be 2D drawings or text files, not an updated version of the 3D model. Attempts to overcome this could be to use neutral formats—for example, STEP files—as both the input and output from variation simulation software. The STEP files can then be read into the CAD system again to update the geometry models. STEP files support PMI but not all GD&T concepts. An MBD approach where no 2D drawings are used, but the 3D models are the basis, removes some of the problems.

 The CAT tool must be able to exchange information with the CAD system (and probably also the PDM/PLM system) regarding:


digital model. To improve "virtual trust", the simulation must be based on correct models and


The information can be exchanged directly or via the PDM system. It is also important to be able to quickly remesh the 3D geometries after changes [37].

The issues related to the simulation working process are also related to the development of the digital model. To improve "virtual trust", the simulation must be based on correct models and reflect reality. Updated 3D models and a clear digital thread are important aspects of this. Moreover, the inspection data used as the input to a digital twin must be of good quality. As stated, the simulation must be seen as a customer of the inspection data. For nonrigid parts, this means that the parts cannot be measured in an overconstrained position, which is usually the case. Instead, they should be measured by locking in only rigid body motions [38] to give a good representation of their actual shape. In that way, the variation simulation can predict the spring-back and the final shape of an assembly. Aspects related to how to choose the inspection points with maximum information content are discussed in [24,39]. Other factors affecting the accuracy of the simulation, and thereby the virtual trust, are listed in [5].

In the future, the interviewees state that a non-nominal digital twin of each individual product is desirable. This is also in line with the results of the survey presented in Section 2. If inspection data is linked to individual products via the digital thread, maintenance can be customized for each individual, which is especially beneficial for high-cost products.

Other aspects given in the list (g)–(k) are related to management and the education system. This is in line with the conclusions in [40]. Those aspects are not core parts of achieving a digital twin, but to use the full potential of a digital twin, they are probably necessary.

#### **8. Conclusions**

To conclude, a digital twin for geometry assurance shows great improvement potential. Examples have shown a reduction of variation on the assembly level with up to 50% compared to a standard joining and assembly process, without individual adjustments and optimization. However, there are barriers that the industry must overcome in order to fully capitalize on those potential improvements. Those barriers are mainly related to the lack of processes for updates and the sharing of the 3D models. Those models must be updated with all changes done during both the development and full production phases. To achieve a complete and reliable digital twin, inspection data must be of high quality and linked to the 3D models. Moreover, universities and research institutions have to elaborate on the underlying digital twin technologies and educational programs for engineers of the future to fully utilize the benefits of a digital twin.

**Author Contributions:** K.W.: conceptualization, methodology, formal analysis, investigation, and writing—original draft; R.S.: conceptualization, investigation, and writing—review; B.S.: investigation and writing—review; and H.W.: investigation and writing—review. All authors have read and agreed to the published version of the manuscript.

**Funding:** The work was carried out at Wingquist Laboratory and the Area of Advance Production at Chalmers within the project "Digital Tvilling", financed by Vinnova. This support is gratefully acknowledged.

**Conflicts of Interest:** The authors declare no conflicts of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

#### **Samad M. E. Sepasgozar**

Faculty of Built Environment, University of New South Wales Sydney, Sydney 2052, Australia; Sepas@unsw.edu.au or samad.sepasgozar@gmail.com

Received: 8 June 2020; Accepted: 3 July 2020; Published: 7 July 2020

**Abstract:** Mixed reality is advancing exponentially in some innovative industries, including manufacturing and aerospace. However, advanced applications of these technologies in architecture, engineering, and construction (AEC) businesses remain nascent. While it is in demand, the use of these technologies in developing the AEC digital pedagogy and for improving professional competence have received little attention. This paper presents a set of five novel digital technologies utilising virtual and augmented reality and digital twin, which adds value to the literature by showing their usefulness in the delivery of construction courses. The project involved designing, developing, and implementing a construction augmented reality (AR), including Piling AR (PAR) and a virtual tunnel boring machine (VTBM) module. The PAR is a smartphone module that presents different elements of a building structure, the footing system, and required equipment for footing construction. VTBM is developed as a multiplayer and avatar-included module for experiencing mechanisms of a tunnel boring machine. The novelty of this project is that it developed innovative immersive construction modules, practices of implementing digital pedagogy, and presenting the capacity of virtual technologies for education. This paper is also highly valuable to educators since it shows how a set of simple to complex technologies can be used for teaching various courses from a distance, either in emergencies such as corona virus disease (COVID-19) or as a part of regular teaching. This paper is a step forward to designing future practices full of virtual education appropriate to the new generation of digitally savvy students.

**Keywords:** virtual reality; augmented reality; digital twin; 360 modules; YouTube; online App; construction; building; digital pedagogy; role play; e-learning; risk management

#### **1. Introduction**

Visiting a real construction site is not always possible due to site restrictions, the limited number of students permitted to enter a site, and, more recently, due to COVID-19. Virtual modules can be used for online education in architecture, engineering, and construction (AEC). They can also be applied to formative learning, flipped classroom [1], blended experimental teaching [2], and online teaching modes [3]. Previous studies investigated the feasibility of using virtual technology in education in different contexts, such as healthcare [4]. Recent studies intend to use virtual technology for education in construction and architecture such as Bashabsheh et al. [5], Wang et al. [6], Eiris Pereira and Gheisari [7], and Gao et al. [8]. However, the application of virtual technologies to show real physical practices in an immersive environment without using headsets for AEC education purposes has not been thoroughly investigated. There are complicated processes in construction, such as drilling and boring underground, which students have not experienced before, and traditional learning methods such as textbooks cannot easily deliver the knowledge. In contrast to text-based learning materials,

there is a possibility of practicing in a simulated environment that allows students for correction and repetition to improve their skills with non-risk failures [9]. The purpose of this study is to present novel tools and online virtual applications to present the complicated processes of drilling, piling, and boring and an excavator digital twin to AEC students. The digital twin refers to the digital replica of a physical entity utilising the internet of things enabling two-way communications between them. The excavator digital twin and other education apps also address the deficiencies of traditional approaches in terms of promoting the engaging capabilities that allow students to be fully immersed in virtual space [6,10,11]. The purpose of these virtual apps is to enable large scale site visits, that will enable students to enter a virtual environment and learn a building case study or heavy equipment. The AEC courses may use project-based learning (PBL) approaches [12], so students are required to enhance their cooperation and collaboration skills [13]. Also, group project learning is recommended to all other educational disciplines at universities [14]. However, several severe challenges of PBL have not been adequately addressed for large classrooms. For example, the way instructors can measure each group member's contribution to their group assignment and give them immediate feedback in large classes will be much more difficult when it comes to the implementation of flipped learning methods or formative learning approaches.

The main research questions are:


#### *1.1. Significance and Advantages of Virtual Technology*

The significance of virtual technologies is that it helps users to have an active experience rather than a passive learning experience and enhances their creativity [15,16]. Emerging technologies and virtual tools have caused a significant change in education methods, including construction education [17]. They shifted education and professional practices significantly away from the traditional individual theory-based lecturing to group PBL, similar to other practical disciplines. Project-based learning refers to learning from a specific construction project as a case study. Examples of group-based learning methods in engineering are problem-solving with open-ended solutions [18,19], hands-on projects [20], and team-oriented communications [21]. More recently, the concept of active learning and student engagement has had a significant impact on education design in practical courses [22], arguing that students learn more and are more prepared for their careers by actively applying the course materials. Some researchers recommended flipped classroom models for construction management [1,23]. However, this new highly lauded method has not yet received enough attention in the AEC, including the construction management discipline, particularly for large classrooms [1,24,25]. Also, there are not enough digital tools to support this teaching method. The problem is that the educator cannot take a large class of students into a construction site. This is particularly an important problem where a specific activity such as piling is not at the same time as the teaching period.

PBL is recommended to many educational disciplines at universities [14]. The PBL assessment is the core concern of many studies in different fields, including AEC, and students need to work collaboratively and enhance their social and cooperation skills. Table 1 presents the positive and negative experiences of students in doing PBL based on the literature.


#### **Table 1.**Overall advantages, experience, and barriers to group project assessments.

#### *1.2. Theoretical Factors*

Virtual education utilizes a set of systems, including hardware and software, that provide an immersive environment or a "sensory illusion" to feel present in a different environment [32]. In virtual education, immersion, learners' perceptions, presence, and interactive activities are known as critical factors. The quality level of these factors, including immersion and interactivity, are related to technological attributes of the utilised technology such as digital images/videos, the display resolution, and other associated gadgets [32]. Lee et al. [33] suggests a set of factors that may increase users' satisfaction of a virtual reality module, including presence, motivation, virtual features, cognitive benefits, usability, reflective thinking. Some factors that are related to the technology acceptance model include usefulness and usability [34–36]. However, some other factors that are related to psychological aspects of learning, such as motivation, may enhance the effectiveness of learning and are usually considered in virtual learning tools. Lee et al. [33] also discuss that the cognitive factors are also important since they enhance understanding and memoriation of the learning subjects in the virtual environment. Memorisation increases the ability of students to recall events, facts, or definitions. Radianti et al. [32] discussed that 'immersion, presence, and interactivity' should be considered in technology design as core characteristics of the virtual module. From a psychological perspective, immersion refers to a state that the student feels isolated from the senses of being in the real world [32].

The literature discusses different factors that can be used for measuring students' experience of using newer technologies and predicting technology acceptance [37,38]. Nakarada-Kordic et al. [39] examined four principal measures of presence, such as feeling 'real', 'relaxed', 'comfortable', and 'anxious', while using the VR, before experiencing magnetic resonance imaging technology in medicine. They found that the VR experience improves participants' experiences before a potentially stressful use of the imaging examination. There are many published papers concerning sickness due to the use of VR applications. Kim et al. [40] developed a motion sickness metric for a successful VR implementation such as discomfort, fatigue, eye strain, difficulty in focus, headache, blurry vision, dizziness, head fullness, and vertigo.

The literature also presents the significance of tools or instructions to help students by enriching learning tasks, activities, and also with relevant experiences [41] (p. 3). However, this might be time-consuming and difficult to help students in large classes. A key component of providing active learning experiences in the classroom is that students need to be prepared for active engagement in the class [42] (p. 369). The flipped classroom is a change in the sequence in which activities are done by which students interact with the course materials [1]. In the flipped classroom, students preview the course materials before class so that they can do a part of their homework and other learning activities in class (workshop) [1]. As explained by Bliemel [43] (p. 113), core lesson materials can be available to students before class, including digital content and short videos. Preparing students before class or supporting them after class one by one is not easily possible in large classes. The purpose of this paper is to offer novel virtual technologies that potentially can be useful for online education and suitable to various class sizes and different teaching philosophies such as flipped classroom, blended learning, role-play learning, formative or summative approaches.

In particular, the current studies also investigate the positive and negative aspects of group project assessment and report that it is not possible to estimate the students' contribution to the main work accurately. However, some other studies provide some strategies to identify free-riders and/or assess other group members' contributions to the group project [44–46]. They offer different methods of peer evaluation or peer-assessments to assist lecturers in identifying the overall contribution of each group member. These methods are mainly based on a simple form asking students to give a mark of 100 to the group members based on their contribution to the group project. This form is an additional source to assess students' contributions to their group projects. However, there is not a reliable tool to detect biases and a valid measure to understand the level of each group member. Therefore, there is a need to address the mentioned problems by developing a novel technology-based model to evaluate group projects' individual marks accurately.

Learning management systems (LMSs) offer useful online tools for managing large classrooms [37,47]. Still, the current systems do not fully support a mix of educational approaches such as roleplay-based group project. In this paper, a set of tools, including a Group Wiki Project (GWiP) is offered as an essential tool for doing an online group project, as shown in Table 2. The GWiP is one of the necessary tools of Web 2.0 that provides spaces to write by students of the group in a web-based setting. This is a constructive activity and constitutes active learning in which students build an individual representation of their knowledge based on their peers' experiences [48].



Männistö et al. [56] reviewed digital collaborative learning practices in nursing education and found out that this type of learning is beneficial since it contributes to knowledge construction and building on each student's interaction. At the same time, they suggest that instructors should provide more guidance to students and design a suitable pedagogical solution by using an appropriate tool for this type of practice [56]. There is a need to develop virtual tools for allowing the opportunity for collaborative learning or cooperative learning. In a collaborative learning process, students are not only working together to carry out their group projects, but they also need to work actively with their group mates and correct each other. Computer-supported collaborative learning systems should be developed further [57]. This type of collaborative learning is known as socio-constructivist [56]. Blau et al. [58] also discuss that this type of practice improves students' self-regulation skills and their learning proficiency by providing peer feedback while completing their project.

This paper first discusses a case study, including materials that should be taught in a construction subject; second, a variety of learning tools produced for online practice will be reviewed; and finally, an overall evaluation of students' feedback is discussed. Limitations and topics for future studies are also discussed in the discussion section. In this investigation, the forefront of contemporary advancements and innovation in AEC education was used to increase the authenticity of learning by being virtually present on site.

#### **2. Research Method**

In order to enhance students learning experience and improve virtual education in construction, this paper developed a set of virtual modules and discussed their applicability in construction. A practical construction course was used as a case study, and details of the course are presented [1]. First, a set of virtual modules have developed, as shown in Figure 1. Then, users tested and used them. The developed technologies were tested by a team of experts, including designers, technical programmers, educational developers, and students during the developing process, as well as before finalising the module. When the development process had been completed, a group of volunteer students was interviewed to learn their experience of using a virtual learning module. In order to improve education technologies, a scientific semi-structured interview was conducted among

construction management students examining their experience and insights gained when using a selected virtual module. The semi-structured interview approach and content analysis are commonly used in the construction context [59–63]. The interviews were analysed manually by using the concept of thematic analysis. Students chosen as users of the modules were also selected to express their experience of using the modules. Their content on GWiP is used as learning evidence. With a focus on the interactive construction reality tour (iCRT) learning experience, the interviews interrogate the potential of the virtual module to support and elevate the students' engagement with the construction process through an immersive or interactive experience. This systematic data collection through interviews allows for a greater understanding of the student perspectives, learning processes, and adoption behaviour [64]. The users' feedback will enhance further development and adoption of VR tools and associated activity-level. Some topics that were asked to be discussed by the interview participants are as follows:


**Figure 1.** Flowchart of mixed reality modules developed for virtual education, including digital twin and online App. VTBM: Virtual Tunnel Boring Machine, PAR: Pilling Augmented Reality, iCRT: interactive reality tour, GWiP: Group WiKi Project. **Figure 1.** Flowchart of mixed reality modules developed for virtual education, including digital twin and online App. VTBM: Virtual Tunnel Boring Machine, PAR: Pilling Augmented Reality, iCRT: interactive reality tour, GWiP: Group WiKi Project.

interactive resources were developed and used for the online delivery of relevant courses in different disciplines. This course was designed and improved by utilising different virtual technologies over five years from the time the course initially was created in 2016. The selected course is called

First-year large foundational core courses. These students have not seen the entire process of

The course was designed based on collaborative learning and authenticity approaches in which

Large classes varying from 200 to 300 students, require developing digital materials for

Cover four learning outcomes to enhance students' ability to understand the processes and

**Table 3.** Selected course learning outcomes for the chosen course about infrastructure construction.

**ID Learning Statement Related Assessment &** 

**Activities** 

Infrastructure and Industrial Construction (IIC), with the following details:

increasing the learning experience while being less expensive;

mechanisms of a variety of construction activities, as listed in Table 3.

unique virtual and online materials will be helpful;

underground activities.

This section presents details of the selected courses, including objectives and topics which

The data collected from ten interviews were analysed in Sections 3 and 4. The students' feedback and their note on GWiP were used for identifying some key factors that may affect virtual technology acceptance.

Figure 1 shows the process, including data collection from construction projects to develop virtual learning modules, including VTBM, PAR, DT, and iCRT. These modules were created for students. GWiP is available on Moodle to students, and a short presentation of their work is available on YouTube. iCRT is available in a cylinder room, namely VR Cinema, and a simplified version is available on YouTube. VTBM is a virtual app that students will be given a link to download the file in conjunction with Discord and Hamachi to use it from home. PAR and DT are available on Google Play or App Store to students so they can download and use it anytime.

This section presents details of the selected courses, including objectives and topics which should be learned by students. The online modules, in this case, were developed to highlight principles that are difficult to recreate for students in the construction and engineering setting. The interactive resources were developed and used for the online delivery of relevant courses in different disciplines. This course was designed and improved by utilising different virtual technologies over five years from the time the course initially was created in 2016. The selected course is called Infrastructure and Industrial Construction (IIC), with the following details:



**Table 3.** Selected course learning outcomes for the chosen course about infrastructure construction.

Table 3 shows the learning statements of the IIC course, which is a first-year core course with 300 students. This course was designed to extend students' knowledge of technologies, systems, and processes of industrial and infrastructure construction. This case study practised this approach by providing a portfolio of activities that includes students (i) forming their role playgroups, (ii) interacting with an industry guest lecturer, (iii) using the immersive environment (VR Cinema) to learn tacit knowledge about a construction case study, and (iv) learning from peers in group projects via GWiP. Arranging site visits for all students poses serious logistical challenges in terms of costs and personal safety considerations. These resources can transform the learning experience of students, very much in line with the objectives of the institutional strategies of improving the learning experience and the education literature. Selected topics covered in this course are shown in Table 4.


**Table 4.** Selected topics covered in the case chosen course.

The GWiP was applied for the Industrial and Infrastructure construction course to give a chance to all students to do their group assignment based on their background and knowledge. The university strategy encourages lecturers to use innovative teaching methods. For a smooth transition between the traditional dominant face-to-face delivery model to the full online classroom model, a combination of different tools was designed and employed, as shown in Figure 1. To increase the authenticity of the modules, a set of engaging industry partners and world-class contractors were involved in the process of developing virtual modules caused in producing useful and valuable sources to students.

#### **3. Technology Design**

This section presents a set of tools and technologies to address the research questions of how mixed reality and digital twins can be applied in construction education and what virtual and augmented reality modules can provide a collaborative environment. These tools were developed for practical construction courses, as shown in Figure 2. Selected participants and stakeholders who are involved in this project include two leading construction contractors and consultants, the university estate management, the portfolio of the pro-vice-chancellor (education), a few instructors, and students from the Faculty of Built Environment, the Faculty of Engineering, and external technology vendors. Students at different education levels were involved in the project. In particular, two undergraduate students, two master's students, and one PhD student participated in the project of recording videos or preparing content for the VTBM module. These participants were involved in creating interactive teaching resources. These resources can be used in different courses such as construction informatics, digital construction, risk management, and practice-based courses, but were mainly used for designing the IIC pedagogy, so-called BLDG1021. Six novel interactive modules have been developed, as shown in Figure 2, and details are provided as follows.

#### *3.1. Group Wiki Project and Role Play*

The first module (refer to Figure 2) is an innovative group project online template, namely GWiP, which was designed for students to do their group projects together, where tutors and other instructors could monitor students' work in real-time. There is a lack of online tools available so far to show the students' progress in real-time transparently. In particular, measuring the student's contribution to their group project is always challenging, but the problem has been solved by the GWiP. All tutors and the instructor use GWiP to monitor students' progress weekly, give them relatively quick feedback, and increase the quality of their group project at the global level. Each group then presented their work (as a scenario base/role play based) and uploaded it on YouTube.

GWiP was perceived as useful technology helping students to prepare their group projects. Rogers [65] suggests that usefulness is one of the two critical factors of technology acceptance and can be used as a construct to predict a successful digital technology implementation. The advantage

of GWiP was perceived as follows: students drafted their project gradually during the semester; instructors, including their tutor and lecturer, monitored their progress during the semester; the 'history' option on GWiP enabled the lecturer to check who contributed the group assignment more than others. GWiP's history page shows the number of words written by everyone with time. recording videos or preparing content for the VTBM module. These participants were involved in creating interactive teaching resources. These resources can be used in different courses such as construction informatics, digital construction, risk management, and practice-based courses, but were mainly used for designing the IIC pedagogy, so-called BLDG1021. Six novel interactive modules have been developed, as shown in Figure 2, and details are provided as follows.

undergraduate students, two master's students, and one PhD student participated in the project of

Appl. Sci. 2020, 10, x 8 of 31

practical construction courses, as shown in Figure 2. Selected participants and stakeholders who are involved in this project include two leading construction contractors and consultants, the university estate management, the portfolio of the pro-vice-chancellor (education), a few instructors, and

**Figure 2.** The flowchart of developed online modules, including screenshots of the applications, including Excavator Digital Twin. **Figure 2.** The flowchart of developed online modules, including screenshots of the applications, including Excavator Digital Twin.

*3.1. Group Wiki Project and Role Play*  The first module (refer to Figure 2) is an innovative group project online template, namely GWiP, which was designed for students to do their group projects together, where tutors and other instructors could monitor students' work in real-time. There is a lack of online tools available so far to show the students' progress in real-time transparently. In particular, measuring the student's contribution to their group project is always challenging, but the problem has been solved by the GWiP. All tutors and the instructor use GWiP to monitor students' progress weekly, give them GWiP increases the transparency of the teamwork, and students do not need to submit or print their projects at the end of the semester since all information will be saved automatically in their computer. For example, a total of 27 groups were formed by students in one semester. Then, 27 GWiP groups were created on WiKi Moodle by the lecturer each semester. Each student was asked to create a page within his GWiP. The group leaders also were encouraged to add extra WiKi pages for their group: 'executive summary,' 'project description,' and 'sharing ideas.'

relatively quick feedback, and increase the quality of their group project at the global level. Each group then presented their work (as a scenario base/role play based) and uploaded it on YouTube. GWiP was perceived as useful technology helping students to prepare their group projects. Rogers [65] suggests that usefulness is one of the two critical factors of technology acceptance and can be used as a construct to predict a successful digital technology implementation. The advantage of GWiP was perceived as follows: students drafted their project gradually during the semester; instructors, including their tutor and lecturer, monitored their progress during the semester; the The students were required to select and analyse a project (e.g., rail work, tunnel, highway, factory, and bridge) based on the relevant information, which they collected. They also were asked to describe their project (on GWiP) in a way that a person not familiar with the project could obtain a clear understanding of the project. The information was related to the course topics, including construction sites (e.g., location and accessibility), construction processes, project organisations, project monitoring, and their learning processes. Groups were encouraged to use theories and topics from the lecture throughout the semester and address them in their project. Based on GWiP, students were able to start their projects in two hours workshop running following the lecture and continue doing their projects any time per week in an online web-based setting.

Subsequently, each student analysed and discussed their chosen project on an individual online page. In this method, they do not use conventional word files, and instead, they described and reported their project on the GWiP page. The online pages allowed the lecturer and tutors to provide regular feedback on their drafts on GWiP. A page called 'Feedback on your Work' was created for each group, and the lecturer and grading tutors use this page to provide feedback to the students of each group. They were asked to identify features of their role, which presents something exciting, challenging, or unique about their role in their real project. They were provided with an example of people involved in the crane operation process, including crane supervisor, coordinator, operator, signaler, and slinger, as shown in Figure 3. It also shows that each group chose one project, and each individual took a

specific role as examples of tasks. Figure 3 presents some of the groups and their workshop ID and the map page of their GWiP, their role, names, and ID numbers. In the first step, students take individual roles, then make the required online pages. They assessed the pages and received group and personal feedback on their work. The GWiP gives the possibility to the lecturer to assess students individually online, and at any time, so each individual within the group is assessable. Appl. Sci. 2020, 10, x 10 of 31

**Figure 3.** Role playgroup project model and list of groups and examples on GWiP. a) Role play Group Wiki Project model; b) Provided sources to students; c) Example of role play scenario for a group; d) List of Groups created by the instructor; f) A selected map page on GWiP. **Figure 3.** Role playgroup project model and list of groups and examples on GWiP. (**a**) Role play Group Wiki Project model; (**b**) Provided sources to students; (**c**) Example of role play scenario for a group; (**d**) List of Groups created by the instructor; (**f**) A selected map page on GWiP.

Then students presented the entire project in a video format and made it available on YouTube (search BLDG1021). Since the course was a large-sized class of over 200 students, applying to the role-play approach and presentation in class was not efficient, so students were asked to use computer visualisation aids and produce their videos and upload them on YouTube. This method helped the student to present the project they have completed during the term. This is in line with the project-based learning (PBL) approach [13]. The resources made by students are always available and helpful to next semester's students. Examples are shown in Figure 4. Appl. Sci. 2020, 10, x 11 of 31 play approach and presentation in class was not efficient, so students were asked to use computer visualisation aids and produce their videos and upload them on YouTube. This method helped the student to present the project they have completed during the term. This is in line with the projectbased learning (PBL) approach [13]. The resources made by students are always available and helpful to next semester's students. Examples are shown in Figure 4.

**Figure 4.** Students' presentations of their role-play practice uploaded on YouTube by themselves. **Figure 4.** Students' presentations of their role-play practice uploaded on YouTube by themselves. cognitive processes that have to be undertaken for a complicated task to be learned. A construction

#### *3.2. Interactive Construction Tour 360 3.2. Interactive Construction Tour 360* site, where students can find themselves surrounded by massive earthworks and equipment worth millions of dollars, can now be experienced first-hand through these new digital and virtual

interactive virtual resources to students.

footages, including some patterns' logo involved in the production.

which depends on the capacity of the cylindrical room as shown in Figure 9.

The second online module (refer to Figure 2) is called iCRT 360, including videos recorded the real construction site of the Science and Engineering Building as a case study as shown in Figure 5. The content of recorded videos was combined with quizzes and interviews with the contractor (i.e., multiplex) on-site staff. The physical area of the immersive iCRT is a stereoscopic and interactive system. It consists of a cylindrical canvas which five high definition projectors visualise the image onto. These five projectors work collectively in sync allowing the 3-dimensional (3D) visualisation of movement to feel smooth and immersive. The second online module (refer to Figure 2) is called iCRT 360, including videos recorded the real construction site of the Science and Engineering Building as a case study as shown in Figure 5. The content of recorded videos was combined with quizzes and interviews with the contractor (i.e., multiplex) on-site staff. The physical area of the immersive iCRT is a stereoscopic and interactive system. It consists of a cylindrical canvas which five high definition projectors visualise the image onto. These five projectors work collectively in sync allowing the 3-dimensional (3D) visualisation of movement to feel smooth and immersive. technologies, in a cost-effective, secure and efficient manner. For more accessibility and further exposure (albeit at lower quality), the 360-degree VR videos were uploaded to a video delivery service that supports VR video, on YouTube. Students could view the VR video content through a desktop/mobile web browser, or through the Android/iOS YouTube apps. If students open these links up in Google Chrome, they can see details of the construction site or pilling process by zooming in/out. For example, if a student makes a mouse scroll, the user will get closer or farther away from the target. Students will be able to click and drag with a mouse to look in all directions of the construction site to explore all around the site.

Additionally, motion tracking systems and stereo sound in the VR environment allowed

The iCRT uses the current technologies and cameras to capture rich information of complex

construction activities for reproducing real practices meaning that students can visit noteworthy events during or any time after the event. The iCRT combines the current experience of Australian **Figure 5.** Panning of an iCRT module sample module used for the case course [66,67](See: https://www.youtube.com/channel/UCOzhGK8xOdoCc3Y9mJHnAnA/ ). **Figure 5.** Panning of an iCRT module sample module used for the case course [66,67] (See: https: //www.youtube.com/channel/UCOzhGK8xOdoCc3Y9mJHnAnA/).

projects and the tacit knowledge of practitioners and uniquely involves them in digital education. The iCRT includes several modules covering underground and excavation activities in construction sites, as shown in Figure 6. For example, one of the modules focused on drilling and the pilling process. These activities are not visible, and students cannot see what is happening underground. Underground activities, excavation, and drilling are not safe for students to visit since there are many hazards in the areas these activities take place. Also, the entire process of piling, including excavation, inserting cages, concreting, may not be possible to carry out within visit times. Since these places are not safe, and the timing is also not under control, there is a need to visualise and create more a b Additionally, motion tracking systems and stereo sound in the VR environment allowed students to interact with the projected modules allowing the opportunity for an immersive experience within an interactive VR environment. Added features in this format were the hotspots that detail specific parts of the construction process, including the equipment involved. Hotspots contained both short videos as well as photos to provide a more in-depth explanation of these processes and equipment. This feature allowed a degree of flexibility by varying emphasis on the description of different aspects of a construction site.

works through cognitive learning in a workplace-based learning environment. Cognitive learning aims to teach learners the processes that experts use to handle complex tasks, situated within the

Figures 7 and 8 show that the module is running using a server computer in the room behind the circular room. However, a tutor can control the module, including backward-forward, pushing the hotspots, and changes the modules using a tablet in front of the students. This is so students can become involved in running and managing the process of using the modules in groups of 10 to 20,

**Figure 7.** The main computer is running the module in a separate control room.

The iCRT uses the current technologies and cameras to capture rich information of complex construction activities for reproducing real practices meaning that students can visit noteworthy events during or any time after the event. The iCRT combines the current experience of Australian projects and the tacit knowledge of practitioners and uniquely involves them in digital education. The iCRT includes several modules covering underground and excavation activities in construction sites, as shown in Figure 6. For example, one of the modules focused on drilling and the pilling process. These activities are not visible, and students cannot see what is happening underground. Underground activities, excavation, and drilling are not safe for students to visit since there are many hazards in the areas these activities take place. Also, the entire process of piling, including excavation, inserting cages, concreting, may not be possible to carry out within visit times. Since these places are not safe, and the timing is also not under control, there is a need to visualise and create more interactive virtual resources to students. look in all directions of the construction site to explore all around the site. **Figure 5.** Panning of an iCRT module sample module used for the case course [66,67](See: https://www.youtube.com/channel/UCOzhGK8xOdoCc3Y9mJHnAnA/ ). Excavator Site

get closer or farther away from the target. Students will be able to click and drag with a mouse to

Appl. Sci. 2020, 10, x 12 of 31

context in which they would usually and naturally be carried out. It also aims to simulate the actual cognitive processes that have to be undertaken for a complicated task to be learned. A construction site, where students can find themselves surrounded by massive earthworks and equipment worth millions of dollars, can now be experienced first-hand through these new digital and virtual

For more accessibility and further exposure (albeit at lower quality), the 360-degree VR videos were uploaded to a video delivery service that supports VR video, on YouTube. Students could view the VR video content through a desktop/mobile web browser, or through the Android/iOS YouTube

technologies, in a cost-effective, secure and efficient manner.

**Figure 6.** Leading a team of internal and external parties for developing the modules from fieldwork to digital lab production. (**a**) fieldwork for production planning in 2017; and (**b**) capturing 360 footages, including some patterns' logo involved in the production. **Figure 6.** Leading a team of internal and external parties for developing the modules from fieldwork to digital lab production. (**a**) fieldwork for production planning in 2017; and (**b**) capturing 360 footages, including some patterns' logo involved in the production.

Figures 7 and 8 show that the module is running using a server computer in the room behind the circular room. However, a tutor can control the module, including backward-forward, pushing the hotspots, and changes the modules using a tablet in front of the students. This is so students can become involved in running and managing the process of using the modules in groups of 10 to 20, which depends on the capacity of the cylindrical room as shown in Figure 9. These resources aim to give the students a virtual experience of significant site and excavation works through cognitive learning in a workplace-based learning environment. Cognitive learning aims to teach learners the processes that experts use to handle complex tasks, situated within the context in which they would usually and naturally be carried out. It also aims to simulate the actual cognitive processes that have to be undertaken for a complicated task to be learned. A construction site, where students can find themselves surrounded by massive earthworks and equipment worth millions of dollars, can now be experienced first-hand through these new digital and virtual technologies, in a cost-effective, secure and efficient manner.

**Figure 7.** The main computer is running the module in a separate control room. For more accessibility and further exposure (albeit at lower quality), the 360-degree VR videos were uploaded to a video delivery service that supports VR video, on YouTube. Students could view the VR video content through a desktop/mobile web browser, or through the Android/iOS YouTube apps. If students open these links up in Google Chrome, they can see details of the construction site or pilling process by zooming in/out. For example, if a student makes a mouse scroll, the user will get closer or farther away from the target. Students will be able to click and drag with a mouse to look in all directions of the construction site to explore all around the site.

Figures 7 and 8 show that the module is running using a server computer in the room behind the circular room. However, a tutor can control the module, including backward-forward, pushing the hotspots, and changes the modules using a tablet in front of the students. This is so students can become involved in running and managing the process of using the modules in groups of 10 to 20, which depends on the capacity of the cylindrical room as shown in Figure 9.

The virtual developed modules have responded to the disadvantage of digital education by dealing with the practical needs of the course. The virtual modules have the potential to change the students' learning attitude to be virtually present on site and increase the authenticity of learning by its virtual presence on-site and being able to see site managers operating in the modules. The modules allowed students to experience real practice, while savings on resources, transportation, time, and money compared with regular site visits made the virtual modules a much more sustainable proposition. **Figure 5.** Panning of an iCRT module sample module used for the case course [66,67](See:

Appl. Sci. 2020, 10, x 12 of 31

context in which they would usually and naturally be carried out. It also aims to simulate the actual cognitive processes that have to be undertaken for a complicated task to be learned. A construction site, where students can find themselves surrounded by massive earthworks and equipment worth millions of dollars, can now be experienced first-hand through these new digital and virtual

For more accessibility and further exposure (albeit at lower quality), the 360-degree VR videos were uploaded to a video delivery service that supports VR video, on YouTube. Students could view the VR video content through a desktop/mobile web browser, or through the Android/iOS YouTube apps. If students open these links up in Google Chrome, they can see details of the construction site or pilling process by zooming in/out. For example, if a student makes a mouse scroll, the user will get closer or farther away from the target. Students will be able to click and drag with a mouse to

Excavator

Site

technologies, in a cost-effective, secure and efficient manner.

look in all directions of the construction site to explore all around the site.

https://www.youtube.com/channel/UCOzhGK8xOdoCc3Y9mJHnAnA/ ).

It also enabled practitioners to optimise their current processes to save more resources. The student's engagement was enhanced using these virtual modules, including iCRT, in the immersive environment. The simple version of my resources can also be used online remotely, which is precisely in-line with the UNSW 2025 Strategic Plan recommending the utilisation of "blended learning products with seamless integration of the physical and digital campuses". These products are critical for enhancing students' learning in construction since construction equipment is costly and project sites are often difficult to access. to digital lab production. (**a**) fieldwork for production planning in 2017; and (**b**) capturing 360 footages, including some patterns' logo involved in the production. Figures 7 and 8 show that the module is running using a server computer in the room behind the circular room. However, a tutor can control the module, including backward-forward, pushing the hotspots, and changes the modules using a tablet in front of the students. This is so students can become involved in running and managing the process of using the modules in groups of 10 to 20, which depends on the capacity of the cylindrical room as shown in Figure 9.

**Figure 6.** Leading a team of internal and external parties for developing the modules from fieldwork

**Figure 7.** The main computer is running the module in a separate control room. **Figure 7.** The main computer is running the module in a separate control room. Appl. Sci. 2020, 10, x 13 of 31 **Rich information srouces**

**Figure 8.** The immersive environment used for the case course at the University of New South Wales (UNSW) (right), A selected Hotspot Site Layout Interview—Bulk excavation (left), the produced module, including the hot spots linked to rich information for each topic. **Figure 8.** The immersive environment used for the case course at the University of New South Wales (UNSW) (**right**), A selected Hotspot Site Layout Interview—Bulk excavation (**left**), the produced module, including the hot spots linked to rich information for each topic. module, including the hot spots linked to rich information for each topic.

**Figure 9.** Three modules, including hotspots in the cylindrical theatre (VR Cinema). The construction site, including hotspots linked to 360 videos available to students. **Figure 9.** Three modules, including hotspots in the cylindrical theatre (VR Cinema). The construction site, including hotspots linked to 360 videos available to students.

#### **Figure 9.** Three modules, including hotspots in the cylindrical theatre (VR Cinema). The construction The virtual developed modules have responded to the disadvantage of digital education by *3.3. Virtual Tunnel Boring Machine*

and project sites are often difficult to access.

*3.3. Virtual Tunnel Boring Machine* 

and project sites are often difficult to access.

*3.3. Virtual Tunnel Boring Machine* 

site, including hotspots linked to 360 videos available to students. The virtual developed modules have responded to the disadvantage of digital education by dealing with the practical needs of the course. The virtual modules have the potential to change the students' learning attitude to be virtually present on site and increase the authenticity of learning by its virtual presence on-site and being able to see site managers operating in the modules. The modules dealing with the practical needs of the course. The virtual modules have the potential to change the students' learning attitude to be virtually present on site and increase the authenticity of learning by its virtual presence on-site and being able to see site managers operating in the modules. The modules allowed students to experience real practice, while savings on resources, transportation, time, and money compared with regular site visits made the virtual modules a much more sustainable proposition. It also enabled practitioners to optimise their current processes to save more resources. The third online module (refer to Figure 2) is called VTBM. It is a game-based virtual environment allowing students to explore how a tunnel boring machine is working underground. This module provides a step by step process involving interactive virtual equipment, where students located in different areas inside or outside the university. They can enter virtually into the VTBM together (Figure 10). The number of students or groups of students allowed into the immersive virtual

The student's engagement was enhanced using these virtual modules, including iCRT, in the

is precisely in-line with the UNSW 2025 Strategic Plan recommending the utilisation of "blended learning products with seamless integration of the physical and digital campuses." These products are critical for enhancing students' learning in construction since construction equipment is costly

The third online module (refer to Figure 2) is called VTBM. It is a game-based virtual environment allowing students to explore how a tunnel boring machine is working underground. This module provides a step by step process involving interactive virtual equipment, where students located in different areas inside or outside the university. They can enter virtually into the VTBM together (Figure 10). The number of students or groups of students allowed into the immersive virtual environment is not limited. Each tutor or student can invite up to ten students to join simultaneously and participate as group members to explore the immersive virtual environment together. The VTBM

The student's engagement was enhanced using these virtual modules, including iCRT, in the immersive environment. The simple version of my resources can also be used online remotely, which is precisely in-line with the UNSW 2025 Strategic Plan recommending the utilisation of "blended learning products with seamless integration of the physical and digital campuses." These products are critical for enhancing students' learning in construction since construction equipment is costly

The third online module (refer to Figure 2) is called VTBM. It is a game-based virtual environment allowing students to explore how a tunnel boring machine is working underground. This module provides a step by step process involving interactive virtual equipment, where students located in different areas inside or outside the university. They can enter virtually into the VTBM together (Figure 10). The number of students or groups of students allowed into the immersive virtual environment is not limited. Each tutor or student can invite up to ten students to join simultaneously and participate as group members to explore the immersive virtual environment together. The VTBM environment is not limited. Each tutor or student can invite up to ten students to join simultaneously and participate as group members to explore the immersive virtual environment together. The VTBM enables them to have the same experience with voice communication available and named avatars for all group members who can then see each other in the VTBM space underground. In VTBM, all students can walk individually through the virtual underground environment and explore different areas of the TBM located underground. In the virtual tour underground, they examine components and tasks relevant to TBM operations such as the cutter head, excavation chamber, mixing arm, bulkhead, screw conveyor, erector, tail skin, tunnel lining, hydraulic cylinders, and the backfilling process. The VTBM is based on a general 3DMax model of a tunnel and some images (e.g., 360 and 3D) collected from different activities so students will be able to explore more realistic underground movements and the TBM operation from any angle using a laptop or a HTC Vive headset. Appl. Sci. 2020, 10, x 14 of 31 enables them to have the same experience with voice communication available and named avatars for all group members who can then see each other in the VTBM space underground. In VTBM, all students can walk individually through the virtual underground environment and explore different areas of the TBM located underground. In the virtual tour underground, they examine components and tasks relevant to TBM operations such as the cutter head, excavation chamber, mixing arm, bulkhead, screw conveyor, erector, tail skin, tunnel lining, hydraulic cylinders, and the backfilling process. The VTBM is based on a general 3DMax model of a tunnel and some images (e.g., 360 and 3D) collected from different activities so students will be able to explore more realistic underground movements and the TBM operation from any angle using a laptop or a HTC Vive headset. Appl. Sci. 2020, 10, x 14 of 31 enables them to have the same experience with voice communication available and named avatars for all group members who can then see each other in the VTBM space underground. In VTBM, all students can walk individually through the virtual underground environment and explore different areas of the TBM located underground. In the virtual tour underground, they examine components and tasks relevant to TBM operations such as the cutter head, excavation chamber, mixing arm, bulkhead, screw conveyor, erector, tail skin, tunnel lining, hydraulic cylinders, and the backfilling process. The VTBM is based on a general 3DMax model of a tunnel and some images (e.g., 360 and 3D) collected from different activities so students will be able to explore more realistic underground movements and the TBM operation from any angle using a laptop or a HTC Vive headset.

**Figure 10.** Screenshots from the login page (**left**), and one of the information hot spots on the VTBM (right). Several hotspots were embedded into the VTBM, referring to some interesting learning points in different parts of the module. **Figure 10.** Screenshots from the login page (**left**), and one of the information hot spots on the VTBM (**right**). Several hotspots were embedded into the VTBM, referring to some interesting learning points in different parts of the module. **Figure 10.** Screenshots from the login page (**left**), and one of the information hot spots on the VTBM (right). Several hotspots were embedded into the VTBM, referring to some interesting learning points in different parts of the module.

The VTBM is a multiplayer operation system module (See Figure 11), which can be used alongside Discord and Hamachi. Hamachi is required when the tutor allows more than one remote student on a different network outside the university to connect as if they are on the same network. Hamachi is a separate application that creates a type of virtual network over the internet. Students can discuss all components while exploring the TBM by using their device's microphone since there is a voice communication option that allows students to communicate with their teammates using Discord. Discord is an optional tool and should be installed separately. This will enable tutors to invite students into channels (potentially one channel per class) and keep the voice conversation continuous and transparent before and after the networked TBM experience is in action. The VTBM is a multiplayer operation system module (See Figure 11), which can be used alongside Discord and Hamachi. Hamachi is required when the tutor allows more than one remote student on a different network outside the university to connect as if they are on the same network. Hamachi is a separate application that creates a type of virtual network over the internet. Students can discuss all components while exploring the TBM by using their device's microphone since there is a voice communication option that allows students to communicate with their teammates using Discord. Discord is an optional tool and should be installed separately. This will enable tutors to invite students into channels (potentially one channel per class) and keep the voice conversation continuous and transparent before and after the networked TBM experience is in action. The VTBM is a multiplayer operation system module (See Figure 11), which can be used alongside Discord and Hamachi. Hamachi is required when the tutor allows more than one remote student on a different network outside the university to connect as if they are on the same network. Hamachi is a separate application that creates a type of virtual network over the internet. Students can discuss all components while exploring the TBM by using their device's microphone since there is a voice communication option that allows students to communicate with their teammates using Discord. Discord is an optional tool and should be installed separately. This will enable tutors to invite students into channels (potentially one channel per class) and keep the voice conversation continuous and transparent before and after the networked TBM experience is in action.

**Figure 11.** Utilising the VTBM on a large screen in an open studying space at FBE (left) and experiencing the VTBM using both a monitor and an HTC Vive Pro headset with another user logging into the VTBM from a different place in a collaborative multiplayer manner (right). **Figure 11.** Utilising the VTBM on a large screen in an open studying space at FBE (left) and experiencing the VTBM using both a monitor and an HTC Vive Pro headset with another user logging into the VTBM from a different place in a collaborative multiplayer manner (right). **Figure 11.** Utilising the VTBM on a large screen in an open studying space at FBE (**left**) and experiencing the VTBM using both a monitor and an HTC Vive Pro headset with another user logging into the VTBM from a different place in a collaborative multiplayer manner (**right**).

In a VR environment, there are 10 ordered drop-in locations, including appropriate thumbnail images, which let students experience ten specific identified sites and read the content provided on the hotspots (See the menu in Figures 9 and 12). Among the hotspots, one represents fresh air, and another refers to an air leak incident describing a fault and its consequences. These hotspots can be useful for risk registration and risk analysis. In a VR environment, there are 10 ordered drop-in locations, including appropriate thumbnail images, which let students experience ten specific identified sites and read the content provided on the hotspots (See the menu in Figures 9 and 12). Among the hotspots, one represents fresh air, and another refers to an air leak incident describing a fault and its consequences. These hotspots can be useful for risk registration and risk analysis. In a VR environment, there are 10 ordered drop-in locations, including appropriate thumbnail images, which let students experience ten specific identified sites and read the content provided on the hotspots (See the menu in Figures 9 and 12). Among the hotspots, one represents fresh air, and another refers to an air leak incident describing a fault and its consequences. These hotspots can be useful for risk registration and risk analysis.

**Figure 12.** Screenshots from the menu and different sections. **Figure 12.** Screenshots from the menu and different sections.

#### *3.4. PAR*

*3.4. PAR*  The fourth online module (refer to Figure 2) is called 'FBE Piling AR' (PAR). FBE refers to the Faculty of Built Environment. The PAR is an interactive virtual environment that goes beyond the traditional pages of a textbook or PowerPoint and enables the foundation construction process to be explained in 4D (3D spatial models plus time). Students will have the ability to play through the animation of the developed typical example of the construction foundation piling work that supports the building. The PAR is an augmented reality app that is available on AppStore or Google Play to all global users. FBE can students download it, and with additional information provided in the course, they will experience the different processes of building construction, particularly piling methods and various types of piling failures. FBE's students can answer relevant quiz questions The fourth online module (refer to Figure 2) is called 'FBE Piling AR' (PAR). FBE refers to the Faculty of Built Environment. The PAR is an interactive virtual environment that goes beyond the traditional pages of a textbook or PowerPoint and enables the foundation construction process to be explained in 4D (3D spatial models plus time). Students will have the ability to play through the animation of the developed typical example of the construction foundation piling work that supports the building. The PAR is an augmented reality app that is available on AppStore or Google Play to all global users. FBE can students download it, and with additional information provided in the course, they will experience the different processes of building construction, particularly piling methods and various types of piling failures. FBE's students can answer relevant quiz questions available on their course webpage on Moodle.

available on their course webpage on Moodle. PAR was developed in two main versions: one version can be downloaded on smart devices (e.g., phones or tablets), and another version is available on Oculus headsets. An unlimited number of students can use PAR. However, if they want to experience the gaming environment as a group, up to ten students can enter into PAR simultaneously. Then they can see avatars in the headset representing their teammates. Avatars can see and ask each other questions when experiencing a different section of the App on the headset version. The avatars' appearance can be edited in Oculus Home, the virtual living room which a student launches into when the student puts on a Rift headset. They can save their experience in the augmented reality environment as a video or image formats, PAR was developed in two main versions: one version can be downloaded on smart devices (e.g., phones or tablets), and another version is available on Oculus headsets. An unlimited number of students can use PAR. However, if they want to experience the gaming environment as a group, up to ten students can enter into PAR simultaneously. Then they can see avatars in the headset representing their teammates. Avatars can see and ask each other questions when experiencing a different section of the App on the headset version. The avatars' appearance can be edited in Oculus Home, the virtual living room which a student launches into when the student puts on a Rift headset. They can save their experience in the augmented reality environment as a video or image formats, either to use the visual material later or to share/communicate their expertise with others/friends on Facebook.

either to use the visual material later or to share/communicate their expertise with others/friends on Facebook. PAR was designed for collaborative, interactive, and engaging practice and includes eight sections, as shown in Figure 1. It represents the construction process undertaken for a multi-story building and provides insights into structural foundation piles for students. The construction process covers site establishment, piling, and constructing the entire structure. The PAR experience is collaborative, i.e., all users see the same model in their VR headsets, allowing exploration and discussion as a group. The collaboration mechanism was enabled using Oculus Quest headsets (Figure 13) connected via a local Wi-Fi network. PAR offers both a mobile-based AR experience and an Oculus headset experience to learn from a construction case study project. In particular, students PAR was designed for collaborative, interactive, and engaging practice and includes eight sections, as shown in Figure 1. It represents the construction process undertaken for a multi-story building and provides insights into structural foundation piles for students. The construction process covers site establishment, piling, and constructing the entire structure. The PAR experience is collaborative, i.e., all users see the same model in their VR headsets, allowing exploration and discussion as a group. The collaboration mechanism was enabled using Oculus Quest headsets (Figure 13) connected via a local Wi-Fi network. PAR offers both a mobile-based AR experience and an Oculus headset experience to learn from a construction case study project. In particular, students can observe different types of failure modes of foundation piles. Students across both platforms can view simplified structures and failure modes of the structural foundation elements (Figure 13).

can observe different types of failure modes of foundation piles. Students across both platforms can view simplified structures and failure modes of the structural foundation elements (Figure 13). The experience shows how the structural foundations of the building can fail due to the quality of the piling. The experience includes a foundation pile construction animation showing the entire construction process, including heavy equipment such as a drilling rig and excavator in different The experience shows how the structural foundations of the building can fail due to the quality of the piling. The experience includes a foundation pile construction animation showing the entire construction process, including heavy equipment such as a drilling rig and excavator in different sections of the virtual animation. This animation gave the student the ability to explore the model

section by section, and students were provided with additional information via hotspots around the model. The PAR model is based on 3D models of the Materials Science & Engineering building and the Kensington campus. The BIM file of that building, campus geographic information system (GIS) data [68] and light detection and ranging (lidar) data providing building height in a 3D context [69,70], were valuable in illustrating the built environment based on a real building. The BIM data available from the case study was used for representing different elements of the building from a pile to a completed building facade. PAR allowed students to look at the structural 'anatomy' of the selected structure (Figure 14) via a mobile device's AR interface or a VR headset (Oculus Quest). sections of the virtual animation. This animation gave the student the ability to explore the model section by section, and students were provided with additional information via hotspots around the model. The PAR model is based on 3D models of the Materials Science & Engineering building and the Kensington campus. The BIM file of that building, campus geographic information system (GIS) data [68] and light detection and ranging (lidar) data providing building height in a 3D context [69,70], were valuable in illustrating the built environment based on a real building. The BIM data available from the case study was used for representing different elements of the building from a pile to a completed building facade. PAR allowed students to look at the structural 'anatomy' of the selected structure (Figure 14) via a mobile device's AR interface or a VR headset (Oculus Quest).

(vi) Multiplayer VR experience of the 3D building model viewed within Oculus Quest headsets on the same local Wi-Fi network; **Figure 14.** The collaborative model experience is presenting the entire structure and piling systems. **Figure 14.** The collaborative model experience is presenting the entire structure and piling systems.

(vii) Visualisation and game experience on students' mobile devices (iPhone 6s or newer and Android S7 or newer); and (viii)Visualisation on Oculus Quest headsets. *3.5. Digital Twin (DT)*  The fifth online module (refer to Figure 2) is an excavator digital twin that is linked to a physical entity of an excavator. The DT module provides a virtual excavator so students can use it to learn different movements of the excavator. This is a step forward toward using a digital twin for education There is an introductory screen in the application to guide the students in how to complete the experience. If there are questions about the visualisation experience, students can ask the instructor or post their queries into a forum created on a learning management system (LMS) such as Moodle. In summary, the PAR offers (Figures 13 and 14):


**Figure 15.** The digital twin AR of the chosen excavator. (**a**) the AR excavator running in the field (https://www.globalconstructionreview.com/innovation/australian-academic-develops-digital-twin-

In order to address the research questions of what factors may enhance students' engagement and identify the advantages of virtual technologies, a group of students were invited to participate. The GWiP documents of 204 students were also screened to identify how students played their roles and how they engaged in their projects. A group of ten students participated in the interviews to discuss their experience of using at least one of these technologies. Students were selected from the undergraduate program of construction management. Students were asked to focus on their roles, which have been taken in an infrastructure project and describe the challenges in their projects. Table 5 shows that they adopted their roles and learned about it. For example, one of a student described her role as a project director: "During my role as Project Director for the Sydney Light Rail Project thus far, many various obstacles have arisen. This is largely due to the size of the project and the high

digger/); (**b**) the digital model on iPad.

**4. Interviews Results and Group WiKi Project Analysis** 


#### *3.5. Digital Twin (DT)*

The fifth online module (refer to Figure 2) is an excavator digital twin that is linked to a physical entity of an excavator. The DT module provides a virtual excavator so students can use it to learn different movements of the excavator. This is a step forward toward using a digital twin for education purposes. The connection between the digital twin and the physical twin required to be on campus, but using the digital version for simulation and students' practice is possible since it was developed for virtual education. This practice may change the educational approach for practice-based courses, as shown in Figure 15.

(**a**) (**b**)

**Figure 15.** The digital twin AR of the chosen excavator. (**a**) the AR excavator running in the field (https://www.globalconstructionreview.com/innovation/australian-academic-develops-digitaltwin-digger/); (**b**) the digital model on iPad.

#### **4. Interviews Results and Group WiKi Project Analysis**

In order to address the research questions of what factors may enhance students' engagement and identify the advantages of virtual technologies, a group of students were invited to participate.

The GWiP documents of 204 students were also screened to identify how students played their roles and how they engaged in their projects. A group of ten students participated in the interviews to discuss their experience of using at least one of these technologies. Students were selected from the undergraduate program of construction management. Students were asked to focus on their roles, which have been taken in an infrastructure project and describe the challenges in their projects. Table 5 shows that they adopted their roles and learned about it. For example, one of a student described her role as a project director: "During my role as Project Director for the Sydney Light Rail Project thus far, many various obstacles have arisen. This is largely due to the size of the project and the high stakes the project holds for those involved in its construction and the impact the infrastructure project would have on the people of Sydney".

The result of the analysis of interviews about learning experience during the module design sessions shows the usefulness of virtual modules as well as limitations, which should be investigated in the future. The value and future research directions are discussed as following. Usefulness is a crucial factor in technology adoption and has been examined in the information systems [71] and construction over the years [64].

The participants support the claim that virtual modules, including the iCRT, provide an innovative immersive environment that brings real construction practices into universities to enable thousands of students from AEC to become familiar with state of the art in a no-risk environment [72,73]. Among these modules, the iCRT was recently experienced by over 1000 students. A student says: "It was more in-depth, and with the questions in the interactive exercises, we got to learn much more than just a YouTube video." Another student expressed the feeling that they visited a real construction site and learned from the visit: "I felt I was on an excursion on-site. It gave me knowledge and insight [into] what actually is involved in a construction site". Table 6 represents a summary of the student experience through the modules.

From the students' perspective, by "becoming immersed" in iCRT modules, a greater sense of "motivation and drive to seek further knowledge on key site elements" has been evident. This includes active engagement within quizzes of the quarry module, where trucks were used to underline the importance of Personal Protective Equipment. The modules provided to students have allowed the cohort to excel in "not just university, but also the workplace" by applying key aspects learned through the modules in the field as a student describes. A student expresses:

"When I first stepped on-site, it was a bit daunting. However, I quickly remembered key concepts of the site layout modules I learned in the VR labs and applied my knowledge to assist in site coordination and asserting the traffic management plan."

Several critical benefits have arisen from the application of virtual reality modules. Students can investigate how "productivity measures" are taken place through 360◦ cameras. An example of this practice includes the number of cycles an excavator takes to fill a truck from the bulk excavation. And through the alteration of crucial selections, the rate of productivity is evident, highlighting the emphasis of efficiency on site, correlating to both cost and time savings. Students have a real feel of the back of house cost and efficiency management, which can be applied on-site first hand in the field. Another benefit is the visualisation of machinery in practice and how events co-occur, highlighting the "concept of critical paths" as a student describes. "We were learning about Gantt charts and how activities have predecessors, and it was great to see how these activities were linked to successfully deliver a section of activities such as the laying of sheet piles." The concept of 'peering beyond the hoardings' highlights the unique experience of VR, producing "a depth of knowledge which cannot be gained unless on-site" one student states. Thus, the module including hotspots with additional and detailed information allowing for greater access to information, which in turn makes students more employable by increasing knowledge and skill base. Table 7 shows the key factors of virtual technologies from students' perspectives. The experience of an immersive environment is hugely enriching compared with just reading a textbook or looking at slides about the construction process.

Table 8 shows that virtual technologies allow for greater insight into construction methodology, expanding exposure to on-site practices. Activities included underground piling, excavating, and mining modules, focusing on the array of different aspects, including operations and risks involved. The virtual technologies directly complemented course content learned in lectures and tutorials where blasting of rock was seen, and shotcrete applied. This provides an insight into construction operations that occur mainly underground and cannot be seen unless on site. The virtual technologies encapsulate students through an innovative medium to transfer vital knowledge about construction, "safety, and scale of operations". This includes "visualising a piling rig's height in comparison to a human on-site with appropriate PPE" as one student describes. "It was great to see the whole operation of the footing system from the piling rigs drilling holes to the reinforcement being in place and finally in situ concrete being poured and capped off to create piles." The depth of knowledge can be seen as students have indicated the positive effect of VR in conveying construction methodologies and procedures.


#### **Table 5.**Summary of results based on the initial assessment and GWiP texts.

Note: IIC stands for industrial and infrastructure construction.


#### **Table 6.**Features identified from the analysis of student experience from selected modules.




**Table 8.**Summary of functionalities and benefits of online teaching tools based on the instructor observation and the literature.

#### **5. Discussion**

This paper presents six innovative tools, including novel virtual digital applications, which specially developed for virtual teaching. The unique virtual technology implementation and students' feedback show that the use of online virtual tools for learning practical construction courses is feasible and useful. This paper describes how selected course content was designed and improved by utilising different virtual technologies over five years from the time the course was created in 2016. Students' satisfactions gradually have been increased, and the course was received firm quotations recently from students saying that they enjoyed and deeply learned the concepts which were not possible to learn from other resources. The use of virtual online technologies potentially enhances students' experience and engagement in practical construction courses. The immersive education modules have promising benefits to students, who are digital natives and known as tech-savvy [76]. These students have an inherent understanding of new online tools, smartphones, and digital devices. The concept of providing an authentic education space with interactive and engaging modules has seen an increase in the depth of knowledge and alignment with course content. Selected benefits were identified as the successful integration of theoretical concepts into practical experience in an authentic learning environment.

This investigation and modules created on virtual education consider the relevance of the online teaching approaches to allied academic disciplines, such as AEC as well as industry-led employee training. It thus speculates as to the online virtual advantages extending from education to economic deficiency and productivity. Visitors from other disciplines were inspired by the modules presented in this paper, including iCRT, and several leading global contractors and educators appealed to them from different universities.

The online versions of PAR and VTBM are significant educational solutions since they help students to obtain knowledge more rapidly and efficiently. The instructor can now spend more time in the online classroom, explaining other relevant concepts related to the visualised components in 3D and thus generate a more practical and complete lesson. This particular example of online practice also creates long-lasting knowledge because students are highly engaged in the space with the components of the building or the tunnelling machinery. Compared with a textbook or a lecture, students understand the practicalities and concepts more rapidly and with greater clarity. The produced modules provide an immersive sense of place when walking through a TBM or a building.

The main benefits of these modules are to improve the learning experience by taking students on a field trip in an immersive environment. In the virtual tour, students will experience a virtual representation of a real project and a simulation of a tunnel boring machine. This exposes students to a practical experience which helps them to understand how a tunnel boring machine works, and to learn the different structural elements of a sample building constructed by a leading construction contractor.

The results show that the online virtual tools are valuable and useful to instructors in construction to enhance students learning and assist them to retrieve their gained knowledge in the lecture in the real context. Future research is required to evaluate students' and tutors' perceptions of the model, using post-implementation qualitative data.

The value of this study is significantly high to AEC educators and scholars in this field who are from older generations teaching digital natives. This paper introduces digital practices to these educators who are accustomed to plan, work, and interact with others in a physical world rather than a virtual environment. The technologies presented in this paper can be considered as a subjective norm for or expectations of the digital natives. They have been using online apps and virtual game-based environments from their early years.

Restrictions on face-to-face teaching and learning due to the COVID-19 pandemic created more demand to use these types of advanced online multiplayer tools to replace construction site visits or laboratory experiences. These modules are designed to be useful for both individual learning and group interactions. They are much more comfortable for individuals to comprehend complex topics compared to textbook reading materials about construction processes. For example, students in a group of up to 10 people could see each other as avatars in a gaming environment and could explore different

areas of the machine, talk about components, and interact with the gaming model. The students experienced the complex components or mechanisms of a machine in VTBM or elements of a building in PAR. The main practical implications of this study are listed as follows: This paper clarified the need and value for designing new online interactive tools for building construction; the practice of effective use of online tools in everyday teaching and learning was discussed; a greater awareness of online interactive tools was made, and relevant design and implementation issues were discussed, and finally a theoretical model was suggested to be examined as future investigation.

Several studies about the mixed reality show significant benefits for students, including improved learning effectiveness (76% more than traditional teaching), engagement, and motivation [77]. By giving students a self-paced interactive virtual learning simulation, students can repeat using the learning materials and experiments without additional costs. The online virtual modules have been designed based on the literature recommendation to include interactive resources such as embedded photos and videos and links to multiple-choice questions. The interactive elements provide different learning scenarios to allow construction students to practice safely and to transfer their knowledge to practice. Where some current commercialised modules simply use expensive digital technologies (e.g., oculus gears and goggles) applicable to small classes, the presented innovative resources were used for massive classes of students. It was not possible to provide oculus gears and goggles to 300 students, and also, it was not possible to take all of them into a construction site. The experimentation showed that the virtual and augmented reality modules gave students a chance to explore building construction processes. They could also experience a tunnel boring machine in operation, identify potential operational risks and hazards (e.g., foundations cracking or leaking issues), and discuss issues while experiencing the immersive environment.

As another example, GWiP is a novel real-time group project model. This model makes the main contributions to the field. First, the GWiP model encouraged students to discuss with piers as each of them takes a role similar to the real projects, and they gave them immediate feedback. Also, tutors could monitor their work in real-time and were able to provide them with feedback when they have any questions. This is in line with the previous recommendation in the literature. For example, Gómez-Pablos et al. [13] discussed that students need help and support for doing their tasks because sometimes they do not know how to work effectively with their group mates or peers. They suggested that their social skills need to be reinforced and improved while doing different tasks at the university [13]. The GWiP helped students to know each other, trusted to peers, and supported each other for constructively accomplishing their assigned works. The role of GWiP model was to allow the opportunity to improve students' interpersonal skills, thereby providing a digital collaboration platform for practising the required skills. The GWiP has also encouraged students to start working on their homework in the workshop under tutors' supervision and continued doing the project outside of class using an online platform under the instructors' supervisor. The GWiP model enabled instructors to examine student's contributions to the group project report in real-time during the semester. However, previous approaches rely on students' judgments that use a different scale of measure.

Based on the interviews, module development experimentation, and the literature, this paper suggests three main factors, including 'perceived usefulness', enjoyment, and engagement as three primary constructs of satisfaction. All these factors can contribute to modelling and predicting virtual technology acceptance. This model is suggested as a conceptual framework that can be examined in different contexts and can be considered as research hypothesis in future studies:

Research question 1: Technology acceptance modelling has a lengthy theoretical background and applied in a different context, but this paper contributes by presenting an extended model applicable to virtual education tools [64,71,78–80]. The article shows some factors that can be considered as measures of "usefulness", which is one of the critical constructs of the technology acceptance model [78,81]. This paper also suggests that 'usefulness' of the virtual technology refers to students' experience of social presence, the possibility of using a rich source of information, and situated learning, which all help students to comprehend detailed practical information of operation process as shown in Figure 16. This is useful since otherwise, it is not possible to obtain the information without involving in a project as a cadet or intern. Figure 16 shows a list of factors identified in interviews and the literature that can be used for modelling and predicting virtual technology acceptance by participants. The proposed model can be examined for different technologies such as DT, VR or web-based virtual technology. The technology can be used for different subjects covering construction operation, risk analysis, safety, and construction informatics. Appl. Sci. 2020, 10, x 24 of 31 identified in interviews and the literature that can be used for modelling and predicting virtual technology acceptance by participants. The proposed model can be examined for different technologies such as DT, VR or web-based virtual technology. The technology can be used for different subjects covering construction operation, risk analysis, safety, and construction informatics.

**Learning technology types**

**Figure 16.** A proposed theoretical virtual technology acceptance model developed based on interviews, for measuring VR acceptance by participants. **Figure 16.** A proposed theoretical virtual technology acceptance model developed based on interviews, for measuring VR acceptance by participants.

Research question 2: The proposed theoretical model also can be modified for 360-degree video applications. For example, the following variables can be examined in modelling: Research question 2: The proposed theoretical model also can be modified for 360-degree video applications. For example, the following variables can be examined in modelling:


Singer presence (WS) measuring student involvement, sensory fidelity, adaptation or immersion, and interface quality [86]. The post media questionnaire (PMQ) is also useful to design a question for assessing the emotions caused by 360 video exposure in terms of "anxiety, disgust, anger, fury, surprise, relax, happiness, and sadness" [85].

Research question 3: How technology-enhanced education and pedagogical design to increase students learning in practical courses, software teaching, and skilled-based subjects in construction. How different elements of virtual online learning will contribute to collaborative thinking and will enhance students' engagement in software teaching and skilled-based subjects. There is a wide range of new technologies such as light detection and ranging (Lidar) tools [69,81,87–90], data mining [91], deep/machine learning [92,93], geographic information systems (GIS) [70,94], different types of laser scanners [88,95–97], virtual reality, three-dimensional printing [98], building information modeling (BIM) [99–102] and digital twin [68,103] that should be practised and learned by students and new employees. The challenging questions concern 'how technology can be used to teach technology'. These tools are often required for employment. Recent education studies recommended that education mode and assessments should be reframed for enhancing students' skills for employability [104]. The questions raise how digital twin and new teaching tools can help educators to 'implement the constructive alignment model' in different subjects [105]. Some relevant research includes the work of Boje et al. [103] and da Motta Gaspar et al. [106] who intended to address this type of question. Still, both education technology and construction digital technologies are advancing, and thus training approaches are required to be modified based on empirical investigations over time.

Research question 4: Some studies tried to use virtual or tangible replicas as a part of digital twin in archaeology using manipulation interfaces (e.g., SketchFab) in archaeology [107]. However, there was no usage of the digital twin in ACE education. Future studies can continue the use of the digital twin in a different context, transferring different concepts to students how the digital twin and a virtual replica can contribute to education transformation and enhance the immersive environment and students' engagements. How can the learning management system (LMS), including Moodle or collaborative Blackboard, as well as communication tools such as Zoom or Microsoft Teams, support these digital simulations. Constructive alignment is challenging when using a digital device. Most of the technologies offer exciting features, but the question is if they will help the instructor to align further the course learning outcome, activities, and assessments. In fact, the question that should be addressed concerns on how virtual technology is useful in implementing the constructive alignment model in teaching.

#### **6. Conclusions**

This project was aimed to present a set of novel technologies and practices, which was used for developing a digital pedagogy. They introduced technologies that enable instructors to monitor students' performance and give them immediate feedback in large classes, where the instructor cannot trace students using traditional approaches. The outcomes and delivered materials discussed in this paper covered different topics such as the 'introduction to a construction project' (e.g., UNSW campus and the selected building) as the first section of the PAR, 'sequences of foundation construction' as the second part of the PAR, and 'design construction processes for tunnelling as a specific activity' covered by VTBM. One of the primary purposes of these online mixed reality modules is to improve the learning experience by allowing students to understand how a tunnel boring machine works, and to become familiar with the different structural elements of a building based on practical experience in a virtual environment. The technologies introduced in this paper offer an opportunity to engage students in acquiring and retaining their knowledge, learn practical knowledge of running an excavator or planning for the excavation process in a construction site as well as improving their teamwork and social skills.

The six primary online education tools were developed and presented in this paper by examining many scenario developments, group discussions, evaluations of initial versions, implementation, and revisions. There was considerable feedback from students to assess each module before the last updated version was finalised. The development team included experts with different backgrounds, lecturers, students, educational development specialists, industry practitioners, and technology experts. Lecturers, students, and educational development experts were involved in designing the modules to underline its usefulness to instructors and learners. Industry practitioners brought first-hand knowledge and case study information into the modules rather than the theoretical information available in textbooks. Details of each education module were discussed in the paper.

This paper is a step forward towards the implementation of a fully online immersive teaching experience in construction, while there are limited practices of the digital twin in the construction education context. The paper presents the potential of online mixed reality in construction education while it has not been thoroughly investigated in line with the advances made in technology development. The recent growth in virtual reality hardware and devices has increased the applicability of mixed reality practices and their ease of use.

The paper also presented the role-play group project model, including individuals and group activities and tasks. The model is implemented on an online platform. The application is beyond 'sharing information in WiKi', and it helps the student to assist each other, give immediate feedback, and correct each other on the report. The implementation of the model has many advantages comparing the current methods discussed, such as:


The contributions of this paper are to extend the body of knowledge in building construction by presenting novel technological applications of digital twin and mixed reality in the construction context. The forms and design development process also present theoretical factors influencing the learning construction process, which increases the learner competency.

The implications of this paper are to present a novel technological approach for building and tunnelling construction education and professional training. Construction project managers can use this approach for two purposes: project induction and training construction operation to novice practitioners. This paper also clarifies the value for designing new online interactive tools for building construction, discusses the effectiveness of online tools in everyday teaching and learning, and raises awareness of online interactive applications in construction education and businesses. This paper presents the outcome of expensive experimentation, which is extremely valuable to construction educators who are from older generations to digital natives. This paper introduces practices to these educators who are accustomed to plan, work, and interact with others in a physical world rather than a virtual environment. However, the technologies presented in this paper can be counted as a subjective norm or expectations of the digital natives. They have been using online apps and virtual game-based environments from their early years. This paper also suggests a plan for future studies, including valuable research issues discussed in the discussion section. The article was limited to presenting potential solutions to the need for virtual reality modules, and the qualitative study limited to examine three modules. Thus, more empirical investigations are required, including surveys, to evaluate each module using a larger group of participants familiar with the virtual apps and tools presented in this paper.

**Funding:** This research received no external funding.

**Acknowledgments:** The practice supported by the Scientia Education Investment Fund (SEIF 2018-2019) and Research Infrastructure Scheme (RIS) at the University of New South Wales, Sydney. With thanks to the participation of instructors, students, technical practitioners, and industry partners and the Construction Mixed Reality Development (CONXR UNSW) stakeholders.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
