There is
similarity and contrast between AR and VR. The ‘Reality–Virtuality continuum’
by Milgram and Colquhoun (1999) shows simply the connection between AR and VR.
They defined Mixed Reality (MR) as containing the entire range of combinations
of virtual and real world components. A subspace of that continuum encompasses
AR. While VR forms a setting where virtual spaces and objects replace the real
world scene entirely, AR forms a setting where virtual spaces and objects are overlapped
on real world scene. This function of AR, that allows users to see virtual
objects and a real world scene collectively, has put it in the limelight of
researchers and developers of various industries (Shin and Dunston, 2008; Meža,
Turk and Dolenc, 2015).

In comparison to VR, AR has greater options of media
representation like ‘text/symbol/indicator, 2D image/video, 3D wireframe, 3D
data, 3D model, and animation’. Each option depicts a class of information with
similar characteristics (Wang et al., 2013).

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

Some studies have tried to simulate construction
activities and get feedback. However, this approach only enables the
visualization of tasks in the virtual setting and roughly relates to the actual
tasks in the real world. AR provides a practical solution to this problem (Chi, Kang and Wang,
2013).

                   
i.           
Related work

Much research has been conducted on the application of
AR in the engineering field. Yabuki and Li (2007) used AR to make a cooperative
reinforcing bar arrangement support system. Wang (2007) utilized AR to guide in
the training of operators of heavy equipment. Golparvar, Pena-Mora and Savarese
(2009)utilized AR to display 4D models used for managing construction tasks.
Schall et al. and Talmaki et al. (2010)  used AR
to display the physical arrangement of underground infrastructure and to lessen
undesired damages. Chi et al., (2012) showed that the integration of AR in
the user interface of construction machinery has a positive impact on crane
operators by developing an AR-Augmented tele-operated crane interface (cited in Wang et al.,
2013).

Progress
in computer interface design and hardware processing power have advanced AR
research prototypes for urban planning (Shen et al. 2001), architectural design
(Kensek et al. 2000), design detailing (Dunston et al. 2002), construction
operations planning (Behzadan and Kamat 2009), execution and inspection of construction
(Webster et al. 1996; Shin and Dunston 2009), maintenance and inspection of
facilities (Navab et al. 2002), revitalization (Donath et al. 2001) etc, (cited
in Wang and Dunston, 2011).

Much of the literature
surrounding AR has taken the form of systematic literature review to synthesize
the current state-of-the-art of Augmented Reality in the construction
industry.

To a greater extent, the AR literature has aimed at the presentation
of visualization and simulation applications for use during
the construction phase that concern to issues faced by on-site workers. With respect to target
audiences, “design team,” “project managers,” and “building systems engineers”
were ‘strong contenders’ (Rankouhi and Waugh, 2012).

Li et al (2018) takes a content analysis-based
review of research and rationalizes the development of VR/AR prototypes in construction
safety, related training and evaluation archetypes. A number of technical
features that could be actualized in the context of construction safety
enhancement are derived. Li et al (2018) also highlights the significant
application domains concerning VR/AR-CS research, viz. hazard identification,
safety training, safety instruction and inspection.

However, they make a weak comparison on how adequately
these bespoke technologies for AR could be utilized to facilitate construction
safety considering the variety of technology characteristics, project types,
scales, work complexity and other factors (Li et al., 2018).

It is noticed that the smallest share of AR research
is ‘training’. It can be justified by the fact that, when speaking of AR, the goal
is to avoid or reduce training and propose a solution which affects directly
the operations (Neges et al., 2017). Through AR, workers could have the
“immediate capability to accomplish the task” (Chimienti et al., 2010;
Palmarini et al., 2018).

                 
ii.           
Technological and
human factors

Shin and Dunston (2008) present that ‘information-intensive’
tasks are notably suitable for support by AR technologies. Also, Webster (1996)  (cited in Wang and Dunston, 2008)
demonstrated that AR systems can be used as field instruction tools.

Inadequacy of information for field operators, inconsistency
between planned solutions and practical applications, and insufficient
communications between related project partners are the main contentions in construction
(Chi, Kang and Wang, 2013).

Most
of the research in AR predominantly falls into two dimensions: horizontal i.e.
the application domain and vertical i.e. the technology domain.  Much headway
has been made on AR technologies such as registration, tracking, and display.
However, a construction AR system must offer usability and be combined with
in-use features to help multi-disciplinary users during the entire construction
lifecycle to achieve industry-wide embracement (Jiao et al., 2013).

Furthermore,
better number of AR systems are designed as independent systems, isolated from
BIM and project management system (PMS). Without adequate integration, the results
of applying AR techniques to BIM/PMS is deficient due to repetitively adjusted
model versions in construction stages (ibid).

In
this study, a video-based, on-line AR system supporting multiple-user
collaboration is submitted. Jiao et al. (2013) shows that this ‘cloud AR’ can apply
real construction data from as-built BIM; by the algorithms an AR scenario can
be designated to multidisciplinary users, and a user can monitor multiple AR
scenarios conceived by disparate users.

The advantages of BIM are due to
its capability to support information management by way of consolidating information.
Its increasing adoption and the related lack of difficulty in data acquisition
has created work settings with information overload, and can thus impact
workers task efficiency otherwise (Chu, Matthews and Love, 2018). Also, the magnitude of effectiveness
of real-time communication in BIM setting is subdued owing to the restricted
sense of immersion into virtual settings. According to Hou and Wang
(2013), AR can be used to accelerate tasks more efficiently as information can
be made available in real-time and real context, thus improving decision-making
(cited in Wang et al., 2014)

Building the logic for the onsite information system, a ‘BIM + AR’
prototypes was devised to show the proof-of-concept of using ‘BIM + AR’
for ‘project control, procurement monitoring and visualization of design during
construction’ (Wang et al., 2014).

Through a design science research approach Chu,
Matthews and Love (2018) developed a ‘mobile BIM AR system (artefact) with
cloud-based storage capacity’. The results showed that the members using the
artefact were 50 percent faster in finishing their tasks, and committed lesser
errors.

To leverage the advantages of integrated BIM and AR
however, a better understanding of the information required by on-site construction
workers is needed. The contextual awareness of AR systems improves the method
of information retrieval by supporting a mechanism to filter data, information
and services, thereby removing redundant data (Chu, Matthews and Love, 2018). Wang
and Dunston (2006) suggest that the feasibility of AR in augmenting construction
related-operations can be evaluated using the following cognitive principles,
viz. information searching and accessing;  attention allocation, working
memory; and spatial cognition (cited in Chu, Matthews
and Love, 2018).
The potential of the MR idea is found in the convenience that are presented for
a user to interact with only the most relevant digital data (Wang and Dunston,
2006; Chi, Kang and Wang, 2013).

However, research efforts must be directed on the
usefulness of AR systems in addition to its feasibility because, situating
information and establishing the connection between the real world and design
information remains the job of humans. (Meža,
Turk and Dolenc, 2015). 

Wang (2005) characterized construction tasks also in
terms of human factors, to make the pertinent connections. This has laid the
groundwork for mapping the MR technologies to specific construction tasks. The technology-mapping
methodology developed, creates the connection between task qualifications and
MR technology elements including information representations and interaction mechanisms,
via human factors analysis through a user-centered approach. (Wang and Dunston,
2008; Shin and Dunston, 2008).

               
iii.           
Application areas

a)     
Maintenance

Palmarini
et al. (2018) assessed the advantages and disadvantages of AR for industrial
maintenance in terms of Key Performance Indicators (KPI) through the results of
a systematic literature review (SLR) on the current state-of-the-art of AR. SLR
concerns with a severe literature review which assures objectivity of the
results. Its use in the maintenance field has revealed several advantages at an
academic level in augmenting human performance in accomplishing technical
maintenance tasks, improving the direction of maintenance operations and backing
maintenance managerial decision making (Palmarini et al., 2018).

 

b)      Progress
monitoring and documentation

In
project management, quick or beforehand detection of actual or potential
schedule delay in field construction activities is crucial. Current progress
monitoring is slow as it needs comprehensive as-planned and as-built data
extraction (Navon and Sacks 2007). Since only visual survey is commonly
performed, the collected data may possibly be subjective. To add to it,
progress reports are visually complicated, and they do not adequately show multivariable
progress information i.e., cost, schedule and performance (Kymell 2008, Poku
and Arditi 2006; Koo and Fischer 2000) (cited in Golparvar, Pena-Mora and Savarese,
2009).

To overcome these disadvantages, this research uses ‘unsorted
daily progress photograph logs’ accessible on any construction site and records
them in the virtual as-planned setting. This provides the unstructured
collection of daily progress photographs to be sorted, interactively browsed
and analyzed. The outcomes of progress comparison between as-planned and
as-built performances are visualized in 4D Augmented Reality setting using a
traffic light metaphor which allows objective progress monitoring (Golparvar,
Pena-Mora and Savarese, 2009).  

On the other hand, Zollmann et al. (2014)
recommends
a method that uses aerial 3-D reconstruction to capture progress information by
suggesting that a staff member taking photographs manually, is comparatively
time consuming and causes areas not being covered very sufficiently. However,
AR must also facilitate supervisors to link possible errors to definite dates
and recognize the accountable workers.

c)      Defect
management

Defect-related-rework, while not adding extra value
also negatively affects the productivity in construction projects that necessitates
additional cost, time, materials, and manpower resources (Park et al., 2013).

The current defect management systems and defect
causation analysis assist in defect correction and also impede the frequency of
the defect, but lack in considering the relationship of defect information flow,
which results in reactive rather than proactive defect management plan (ibid).

In this study, Park et al. (2013) applied a marker-based
AR and image-matching techniques to the defect control on a construction site.
Using these AR techniques, supervisors and on-site workers could instinctively verify
the outcome of their tasks by augmenting virtual shapes and dimensions onto the
real objects. AR can allow defects due to omission error and dimension error to
be controlled. The field application potentials of the AR and object-matching
inspection techniques have well confirmed through the lab experiment (ibid).

d)      Site
layout planning

Current
future worksite layout planning is greatly prone to errors because it relies 2-dimensional
paper modeling which does not support the spatial sense of the users. Also, the
spatial 3D view of the entire worksite has to be mentally constructed in the
planners mind which results in improper planning logic. The AR has been
proposed as a technique to reasonably prevent potential planning errors and
process inefficiencies (Wang, 2007).

AR ‘Planner’ assists the construction worksite
planner to place virtual 3D models of construction materials and handling
devices in a real world scene and chart the corresponding routing lines in the
planned worksite. Users could maneuver the virtual objects interactively to
design and plan the configuration of the construction worksite. This technique
also supports animations of the static objects to capture the behavior of the
dynamic components (ibid).

e)      Training

It is envisaged that the state-of-the-art in
training can be advanced with AR. The current method of equipment operator
training, either by off-site training or VR gives a novice narrow opportunity
to experience the real working environment (Wang and Dunston, 2007).

Wang and Dunston (2007) show the advantage of an
AR-based real world Training System (ARTS). For example, ‘a novice operator
sitting in a real front end loader cabin can manipulate the equipment to move a
virtual soil stockpile on the ground to a dumping container certain distance
away.’

Some studies demonstrate the reliable transfer of
skills acquired through virtual training programs to substantive on-the-job operations,
provided a high fidelity between the virtual and real settings (Rose et al.,
2000). ARTS could expedite the progress along a steep learning curve (ibid). Such
real-time and real-world training systems could assist novice operators through
complex construction tasks (Feiner, Webster and MacIntyre, no date).

f)       Landscape
preservation

High
rise construction can destroy the landscape of other important buildings from
multiple viewpoints. To check if some portions of the high rise buildings are in
sight or not, behind the ‘locations of interest’ from multiple viewpoints, it
is essential to make a 3D model representative of the geographical features,
existing structures and natural objects. Yabuki, Miyashita and Fukuda (2011)
show a method using AR, where a number of 3D virtual rectangular objects with a
scale are located on the grid of 3D geographical model. Measurement of the
maximum invisible height for each rectangular object at a grid point can be
made by overlapping the virtual objects with the actual landscape from multiple
viewpoints. This method was verified for it feasibility and effectiveness by a
design-research approach (Yabuki, Miyashita and Fukuda, 2011).

               
iv.           
Barriers to
implementation

First, the question of how this technology integrates
into the information technology infrastructure of the company needs to be
solved (McKinsey & Company, 2016).

An
important issue is the current state of the technologies that support AR.
Currently, these technologies are not mature enough to be effectively applied
on real construction project sites. For example, tracking technology which
aligns the user’s viewing direction and position is an important concern for precise
registration of virtual objects in a real world scene (Shin and Dunston, 2008).

Presently, AR systems are tested in a
strictly-controlled setting which assures accurate tracking, which does not
assure adaptability in un-controlled environments. A powerful AR system must
work in all settings without having to learn these settings in advance (Wang,
2009).

Another
challenge for developers while designing AR systems is the extraction of
industrial domain knowledge . This will require a systematically organized and integrated
3D database that could be readily applied by AR systems. It is generally
observed that AR systems tend to exist on non-standardized hardware and
software which impedes its widespread use and integration. (Chu, Matthews and
Love, 2018; Palmarini et al., 2018).

Lastly,
many reported AR systems use devices that are inconvenient, causing ergonomics
issues such as fatigue and discomfort (Jiao et al., 2013). A well developed
natural user interface (NUI) provides more comfortable and perceptive user
experience, which can raise the usability of AR (Chi, Kang and Wang, 2013).

 

Chapter
5. Research methodology

Creswell
(2014) defines research approach as “Plans and procedures for research that
span the steps from broad assumptions to detailed methods of data collection
analysis, and interpretation.”
(cited
in Haddadi et al., 2017).

                               
i.           
Research strategy

The choice of research
strategy will provide data which will pave the way for the exploration of the
research objectives. The two types of research 
strategy are qualitative and quantitative (Fellows and Liu, 2008; Bell,
2010; Naoum, 2007).

Quantitative research proceeds with analysis of the
relationship between one set of data and another.

Qualitative approach is principally connected to the
interpretivist view of philosophy (McLaughlin, 2012). The qualitative method seeks
to gain insight and to comprehend the user’s viewpoint of the world (Fellows
and Liu, 2008) Interviews, case studies and literature reviews are the usual
methods for data collection in qualitative studies.

In a qualitative approach the analysis of the data
builds inductively from particulars to common ideas and finally the researchers
make interpretation of the data and it is considered to be the best approach to
seek the research objectives of this study.

A phenomenological method of conducting the
qualitative research is proposed, as it will help in evaluating the experiences
of individuals about a phenomenon (Haddadi et al., 2017).

Because every construction project is built in different
settings, the expectations and experiences of the project stakeholders will
differ in relation to AR. To accommodate the differing views of the potential
and existing user’s of AR, the interpretivist approach is most suitable for
this study.

                             
ii.           
Research design

 “Research
design logically links the research questions to the research conclusions
through the steps undertaken during data collection and data analysis.” (Baškarada, 2014) (cited in Twining et al., 2017; Haddadi et al.,
2017).

In a cross-sectional design, also known as the
survey approach data is collected from a number of participants at the same
point in time (Neuman, 2014). According to Bell (2010), there is no single
definition of ‘survey’, but it is concerned with the opinions and attitudes of
people. As cross-sectional design approach allows causal inferences to be made,  this method is most suitable to answer the
research question of how AR can affect the time, cost and quality constraints
in a construction project.

                           
iii.           
Research method

Because AR is somewhat unfamiliar; it is necessary
to choose a research method, which allows a definite presentation of the topic.

To assure the validity of the collected data, the
sample selected will only consist of those people who are able to understand the
typical building documents, like project managers, architects, engineering
consultants, developers, etc. Thus, purposive sampling is suggested for this
research (Wang et al., 2013).

The research objectives require the
insights from the perspective of end-users. Semi-structured interviews  are proposed to collect appropriate qualitative
data as it provides the flexibility of an open-ended interview and the advantages
of a structured ethnographic study. This method can reveal good quality descriptive data
on the individual experiences of the participants.

A pilot study is envisaged to ascertain the quality of
the questions, correct errors, omissions or logical inconsistencies. The
questions will majorly
be open ended because the issue is not a factual and can have a wide range of
user opinions.

Content analysis is a research
technique used to reproduce and validate inferences by coding textual
material. Content analysis is beneficial as it allows to examine the user
perceptions. The qualitative data
generated will be screened for patterns and coded into categories to be displayed
in the form of bar graphs, line graphs and pie-charts by grouping the
categorical data into non-overlapping categories. As far as possible, responses
that ask for user perceptions and attitudes should be quantified and described
in text (McCammon, no date).

The ethical issues regarding respecting the rights
of respondents, such as anonymity and data protection will be assured (O’Brien
et al., 2014; Twining et al., 2017).