A COMPARATIVE ANALYSIS OF ARCHITECTURAL EXPERIENCE IN

THE REALITY CONTINUUM BASED ON EMBODIMENT, TACTILITY

AND AFFORDANCE

AVISHEK DAS1, LARS BRORSON FICH2 and CLAUS

BROENDGAARD MADSEN3
1,2,3
Department of Architecture, Design & Media Technology, Aalborg
University
1
[email protected], 0000-0001-8386-8339
2
[email protected], 0000-0002-6760-0540
3
[email protected], 0000-0003-0762-3713

Abstract. This paper presents a comparative analysis between a

physical model, a set of virtual reality (VR) models and a set of
augmented reality (AR) models of the same architectural project. A
set of architectural participants’ interactions with the models have
been recorded through a set of sensor streams and videos and their
experiences have been interpreted through a questionnaire containing
both qualitative and quantitative questions. Further analysing through
the sensor streams, video recording and questionnaire, their
experiences have been interpreted using a framework of embodiment,
affordance and tactility to determine the usefulness and limitations
and each of the modes and their possible application in architectural
design practice.

Keywords. Augmented Reality, Virtual Reality, Architectural Model,

Embodied Cognition, Affordance, Spatial Cognition

1. Introduction
Historically, in the architectural design process, two-dimensional drawings, physical
models, and materials have played a crucial role in visualising and experiencing the
space before it was built(Mindrup & Wells, 2020). The role played by the physical
model in this paradigm was important and preferred by many artists, architects, and
patrons throughout history for various structural and experiential properties.
Architects like Callimachus of Greece during the Hellenic period, Michelangelo and
the Medicis during the Renaissance, Antoni Gaudi during art nouveau, Frei Otto
during Modernism and Alvar Alto, Peter Zumthor, Rem Koolhas in the post-modern
to present day have explored the physical materiality and spatial qualities with
physical models of varying typology and scale. In the past 30 years, with the
development of technology, computer-aided design (CAD), 3D modelling
capabilities, building information modelling (BIM) and photorealistic visualizations
have grown and have developed on an unprecedented scale. Consequently, these

HUMAN-CENTRIC, Proceedings of the 28th International Conference of the Association for Computer-Aided
Architectural
– LEAVEDesign Research
THIS WHITE BOXinON
Asia (CAADRIA)
PAGE01!!– If it has2023,
moved,Volume 2,and
you can cut 69-78.
paste it©back
2023 and1,published
to page right click onbythethe
Association for Computer-Aided Architectural Design Research in Asia (CAADRIA), Hong Kong.
boundary and choose 'More Layout Options...' and then under 'Vertical', choose 'Absolute position' - 24 cm (below Page).
70 A. DAS, L. B. FICH AND C. B. MADSEN

developments have majorly influenced the architectural design process in many

ways. The efficiency and ease of 3D CAD and BIM models have many ways
replaced the practice of using a physical model in many architectural practices.
Additionally, using immersive virtual reality (IVR) many architectural practices have
almost entirely replaced the architectural model except for the final presentation stage
from a material, cost and time involvement perspective while still obtaining valuable
spatial and architectural insights(AIA, 2006). The physical and virtual reality models
reside on two extremes in the reality continuum(Fig. 1) (Milgram & Kishino, 1994).
Hence, the question arises if these VR models can entirely replace the role of physical
models or if different modes of models can still contribute anything specific to the
design process.

Figure 1:Reality Continuum, Drawn after Milgram & Kishino 1994

Previous studies have been conducted by Sun et al to determine if the spatial

reasoning capabilities are similar between a computer graphics model seen on a
screen and a physical model of the same (Sun et al., 2013). The study almost entirely
focused on estimating dimensional perception between the models but did not offer
much insight into issues of embodiment and tactile perception as the participants
were not allowed to touch the models. Though the study concluded by mentioning
that the physical model was better at height and distance estimation, it is imperative to
mention that the other setup was not an immersive virtual reality but a computer
screen-based virtual reality where movements were restricted. Later in 2018,
Hermund et al studied the perception of an architectural atmosphere through VR,
physical space itself and 2D representation of the space through plan and section
(Hermund et al., 2018). In their study, VR simulates similar nature of perception of
being in the actual space while 2D drawings like plans and sections and photorealistic
non-immersive visualizations are not as effective as the former one.
In the past 4 years due to the rapid development of inside-out tracking technology,
consumer-grade standalone augmented reality (AR) and virtual reality (VR) devices
have become available which can be used in a smaller space through internal
tracking. Similarly, ARKit and ARCore for iOS and Android devices have brought
AR to mobile devices democratizing these mixed reality (MR) applications. Also, in
physical manufacturing, rapid prototyping technologies have reduced the fabrication
time for 3D objects. A combination of these technological advancements has helped
to explore the different modalities along the reality continuum in a much-simplified
manner.
In this project, we have developed four basic modalities of one architectural
model: a physical model; through an immersive VR headset (Meta Quest pro); a
handheld AR device (iPad Pro) and through a head-mounted AR headset
 A COMPARATIVE ANALYSIS OF ARCHITECTURAL 71
EXPERIENCE IN THE REALITY CONTINUUM BASED
ON EMBODIMENT, TACTILITY AND AFFORDANCE.

(HoloLens2). Participants with architectural education were selected to interact with

the model with a follow-up questionnaire about the spatial configuration of the
architectural model they have interacted with. Through both qualitative and
quantitative analysis of the data obtained from the participants, the study will
fundamentally investigate two key interests: (1) how different degrees of embodiment
affect the scale and spatial perception of the space through different modalities and
(2) how affordances of these modalities influence the interaction with the model and
in turn how it affects our perception of the space. Based on our research interest we
would like to put forward the hypothesis: Spatial perception offered by the different
modalities over the reality continuum are different in nature and should not be ideally
interchanged for the sake of ease of use. In this paper, we will present the effect of
embodiment, affordances, and tactility on the understanding of the spatial perception
and spatial cognition of the user by studying the participants’ degree and manner of
interaction with the model. Finally, the paper will identify the unique role of each
modality over the human perception of architectural space.

2. Methods
French Philosopher Maurice Merleau-Ponty, in his seminal work, represented the
“body” not only from the sole anatomical perspective or a static passive entity but as
an entry point to explore reality; the body as a tool with certain affordances that helps
us experience our immediate world(Hale, 2017). In this case, the body acts as a
compass of navigation in our immediate world, be it virtual or real. On a very similar
note, for the ocular-centric approach to architectural experience, renaissance master
Michelangelo often used the metaphor “seste negli occhi” or compass in the eyes
emphasizing the importance of appearance and the degree of realism that
encompassed architectural representation(Conforti et al., 2020). Through these ideas
“body as a compass” and “eye as a compass” we would like to like to explore the
importance of movement, visual quality, and embodiment in spatial cognition.
In this experiment, participants needed to experiment with a single architectural
project in different modalities. An architectural project was chosen among three
shortlisted projects showing enough spatial complexity and consisting of volumes of
different scales. In this case, it is a model of an institutional building made for a
cultural icon. Physical and mesh models were made of the same in 1:200 scale of
similar material appearance. In this case, the material was chosen to be white material
without any resemblance to any physical material that might have been used on the
building. Regarding the level of detail, this model could be classified as a LOD 300
model.

2.1. PARTICIPANT SELECTION

This experiment aims to implement the results in the discourse architectural design
practice. Thus, the participants chosen for the experiment have a minimum of five
years of architectural training in academia and a basic understanding of mixed-reality
tools. The experience of the subjects varies from two years after graduation to twenty-
five years of experience. The participants Also, in terms of experience with MR tools,
there is a good mix of amateurs to experts among the participants. All the participants
72 A. DAS, L. B. FICH AND C. B. MADSEN

are from academia with a range of research assistants, assistant and associate
professors and PhD fellows.

2.2. ARCHITECTURAL PROJECT

The architectural project used in this experiment is an unbuilt competition entry of a
memorial for a cultural icon from Square Consultancy Services, Kolkata. The model
has been used with their permission. The project can be classified as a cultural
building which encompasses a stairway to the main floor and a spiralling volume
with large balconies towards an artificial water body.

2.3. PHYSICAL MODEL

The physical model was made using laser-cut white cardboard and 3D-printed parts.
The laser cut burns were kept in the model as a distinct nature rather than covered up
through paint to retain the appearance of an architectural model. The model was also
fitted with an inertial measurement unit and a data logger to record the movements in
terms of rotation around the X, Y and Z axis while handled during the experiments.
The physical model has a dimension of 420mm x 400mm x 60 mm on a 1:200 scale.
It was also fitted with an HTC Vive Tracker which helped to corroborate the IMU
data using external tracking.

2.4. VR & AR APPLICATION

A CAD model was made in Rhino3D which was consequently converted to a closed-
mesh model to export in VR and AR applications. For the VR application, Meta
Quest Pro was used in the experiment as it became available during the process and
the colour pass-through mode helped decide a certain aspect of the experiments.
Arkio (Arkio - Collaborative Spatial Design, 2022;) was used as the VR application
to facilitate the viewing of the model in 1:1 and 1:200 scale. Arkio also offers a
passthrough mode through which users can see the model in their surroundings,
mimicking augmented reality while retaining the same graphic quality.
AR application was divided into two subcategories: handheld and head-mounted.
For handheld AR, an iPad Pro with a 13-inch screen was used while Hololens2 was
used for the head-mounted option. For both cases, Fologram(Fologram, 2021) was
used to connect and facilitate the AR experience as well as track users’ movement.
The choice of iPad was added to the mix due to the price point of the Hololens2. To
avoid discrepancy in experience it was decided to advise the participants not to move
the AR model as the mode of interaction was different in two devices: one was
through the touchscreen, and another was through spatial grabbing/pinching using
hands.
During the experiment, a set of sensor streams was captured from the physical
model and head-mounted displays and iPad. After the experiment, each participant
had to answer a questionnaire about their experience with the models in different
modalities. In this experiment, only commercially available and consumer-grade
hardware, software, and applications were used to collect data from the participants.
 A COMPARATIVE ANALYSIS OF ARCHITECTURAL 73
EXPERIENCE IN THE REALITY CONTINUUM BASED
ON EMBODIMENT, TACTILITY AND AFFORDANCE.

3. Design Experimentation
Though the visual perception of the space appears as the primary stimuli or the
gateway to experience the architecture it is a multisensory and tactile
discourse(Pallasmaa, 2012). Spatial volumes alone can evoke certain emotions but
our bodily interactions like walking, touching, and smelling can append/modify our
base understanding of the space. This sensory-motor coupling helps us create a better
understanding of the world(Hostetter & Alibali, 2008). Often through computer
graphics and visualization, we tend to compress this multisensory experience into an
ocular-centric experience which is essentially the case of immersive virtual reality or
augmented reality-based representation.
In this design experiment, the participants have experienced one space in different
modalities which are spread over the reality continuum encompassing a different
level of embodied cognition. Each participant in this experiment went through 6
different modes of representations of the space: physical model in 1:200 scale, VR
model in 1:200 scale, 3D models in 1:200 scale with passthrough enabled which has
the controls and visual qualities of Arkio but seems like an AR, VR model in 1:1
scale, AR model through Hololens2 and AR model through an iPad. For each mode
participants got two minutes to explore the model once they became comfortable with
navigation in that mode. In each participant’s case, the order of the modalities was
shuffled to ensure a non-biased opinion on the experience itself. To ensure the
participant can navigate the model space, three control questions were asked at each
stage; if they can clearly see the model, if they can navigate with ease and if they feel
at ease with the modality and the environment; once they were comfortable the
experiment continued. The experiment was conducted in a studio space where
students usually work with physical models and projects with which the participants
are familiar. This choice was made to ensure the experiment space should not hinder
the movement of the participants. After each participant went through all six
modalities, a questionnaire was presented to the participants to record their
experience which had both qualitative and quantitative parameters.

4. Results
The results of the experiment gave insight into how different modalities can offer
different experiences on the same model. According to the post-experiment
questionnaire, 71% of preferred the 1:200 scale model as the most favourable
exploratory medium while 29% preferred the VR 1:1 model. But as a second
preference, 66% chose VR 1:1 model. On the other hand, 71% of participants ranked
74 A. DAS, L. B. FICH AND C. B. MADSEN

Hololens2-based AR as the least preferred option and 29% mentioned iPad-based AR

owing to mostly their translucent and “fragile” style of visual representation (see Fig.
2). Further insights are classified into three categories: (1) movement in and around
the model, (2) embodied cognition/bodily perception and (3) scale perception.

Figure 2: Modality preference and Scale perception

4.1. MOVEMENT & EMBODIED COGNITION

For all the modalities, the headsets and handheld devices were tracked through either
inside-out tracking data or outside-in tracking. The finding is explained in the
following sections.
When we look at a 3D object naturally, we hold it in our hand or move around the
object based on the affordances. In the case of the physical model, it has been
observed that everyone picked up the model and mostly rotated the model around the
Z axis. In comparison, the movement around the X and Y axis is quite minimal (Fig.
3). This data is mainly obtained from the IMU fitted with the model and the tracked
path through HTC Tracker is also in tandem with this observation.

Figure 3: Rotation Data Distribution of Physical Model

 A COMPARATIVE ANALYSIS OF ARCHITECTURAL 75
EXPERIENCE IN THE REALITY CONTINUUM BASED
ON EMBODIMENT, TACTILITY AND AFFORDANCE.

A similar observation was found in the case of AR-based visualizations but in this
case, participants chose to walk around the model as picking up in the model was not
a viable solution, but the patterns of the movement are different in nature. In the case
of Hololens2-based AR, participants mostly walked around in mostly circular pattern
while looking at the model from a distance (Fig. 4).

Figure 4: User Movement tracking through AR Devices

Figure 5: User Perspective and Extent of User reach in AR

The narrow field of view of Hololens2 could be a contributing factor to this

behaviour. In contrast to this observation, in iPad-based AR, participants moved more
freely while looking at the model and due to the affordance, they could look inside
the model (See Fig. 5). The handheld nature of the device gave the participants a
unique perspective on the model and the ability to control the movement. Also,
participants reported, the attachment to the device made them less aware of their
bodies and the affordance of the iPad extended to the body which limited the body
76 A. DAS, L. B. FICH AND C. B. MADSEN

movement.
In the case of VR 1:200, the observation was like the physical model; participants
picked up the model and put it up at eye level while rotating the model itself. But in
the passthrough mode, participants were observed to move around more around the
model as they were made aware of their surroundings (See Fig. 6). This phenomenon
happened despite the participants knowing they were within a safe boundary, and
they could be warned if they wanted to leave the same.
But in the case of VR 1:1, participants observed some difficulties as they could
not walk around a large model or climb up a stair in the model. In the 1:1 mode, the
participants need to “teleport” themselves in the model using the joystick, which
resembles jumping movement in the real world and deters the participants from
exploring the model compared to the other modes. Often in this mode participants
found themselves lost in the model. This effect is known to be present in the VR and
causes disorientation of the participant in the space as the participant is not using their
body to move around causing a disjunction between sensory and motor
function(Prithul et al., 2021).
Figure 6: User Movement in VR & Passthrough

4.2. SCALE PERCEPTION

As humans, we usually do not look at one object in isolation but create a set of
relationships with the object and its immediate surroundings(Berger, 2008).
Participants were asked if we consider the 1:200 physical model as a benchmark then
which modality was the closest in scale compared with the physical model. Here
again, 72% of the participants mentioned VR 1:200 in passthrough mode was the
closest in the scale while 14% Hololens2 visualization was the most similar in scale
while the rest of the participants mentioned that iPad AR was closest in scale
comparison. Their close relationship with the surroundings in those modes could be a
driving factor for that result. In VR 1:1, participants’ scale perception in one single
perspective is reported to be well while understanding spatial configuration is
reported not well due to the limitation in movement. This could stem from
participants’ inability to map the space using their own stride which could have given
a mental map of the space. Also, in the case of the physical model, the tactile
perception and grabbing the model with their own body reportedly gave the
participants a better understanding of the scale which was totally absent in the AR-
based visualizations but faintly present in VR-based ones due to the VR controller.
 A COMPARATIVE ANALYSIS OF ARCHITECTURAL 77
EXPERIENCE IN THE REALITY CONTINUUM BASED
ON EMBODIMENT, TACTILITY AND AFFORDANCE.

5. Discussion
In this experiment, it has firmly been ascertained that each of the modalities has its
benefits and limitations. The agency of the physical model is well established in the
architectural educational discourse which helps architects lean towards the physical
model given a choice. But VR is being used in most practices for its versatility in
scales and advantages over quick visualization, like how rapid prototyping disrupted
the manufacturing industry. But the limitation of movements in the VR environment
even in the commercially available tools sometimes hinders the purpose of the tool.
Due to technical development in the last two years, VR headsets have leapt forward
towards modes like passthrough which allows the user to the ground reality while
seeing the model in the device. These modes enable the user to be “present” their
reality and still view the models in another reality thus ranking higher on embodied
cognition. But again, the lack of haptic interaction with the model makes the physical
model much more viable for a quick study without the technological overload.
Though it has been found that AR is not being used in architectural offices(Das et
al., 2022) as actively as VR and physical models, their utility through low-cost
handheld devices could offer certain pertinent insight. These devices can be used to
view the model from the outside as well as can help to look inside the model
exposing the spatial configurations. But the visualization style for the AR
visualizations needs to be improved by a certain degree to make it a viable
alternative.
Based on the evidence presented it could be indicated that due to classical
architectural training, many architects might prefer the physical model but to explore
the space in detail, AR models could pose a significant opportunity for architects
using low-cost handheld devices. On the other hand, through VR one can study the
volumetric quality of the space better than a physical model or a screen. These
technologies are part of a spectrum, and their utility should be complemented by one
another based on the goals and relevant stages of architectural design, pertinent to the
advantages and affordances they provide.
Some technological development is focused on developing haptic interaction
devices for VR but another intermediate stage between the user and the object could
again create the distance between the object and the user. But similarly, if these
realities could overlap and inform one another in the architectural discourse, e.g., data
visualization over the physical model like pre-visualization of unbuilt works on-site,
hypothetically the acceptance of the technology could be increased by addressing the
value creation and haptic capability. In all these studies, material properties and
material representation play a lesser role due to technological limitations and an
understanding of the digital representation of materials. The graphical representation
of the models in computer graphics always reminds of a “finished” object while some
physical materials like cardboard or foam give a sense of “unfinished”(Rowe &
Bechthold, 2017). If this material representation can be obtained in mixed reality,
then the acceptance of the technology could be arguably higher.

Acknowledgements
The authors would like to thank all the participants for their time and involvement in
78 A. DAS, L. B. FICH AND C. B. MADSEN

the project; the CREATE workshop team who helped to create the models lent their
expertise in fabrication and a special thanks to Post-Doctoral Fellow Kasper Hald for
his help in establishing HTC Vive Tracking. We are also thankful to Square
Consultancy services who have generously allowed their model to be used in this
experiment.

References
AIA. (2006). Report on Integrated Practice.
Arkio - Collaborative Spatial Design. (2022). https://www.arkio.is/
Berger, J. (2008). Ways of seeing (Issue 1). Penguin.
Conforti, C., Colonnese, F., D’Amelio, M. G., & Grieco, L. (2020). The Critical Agency of
Full-Size Models, from Michelangelo and Bernini to the Picturesque Garden.
Architectural Theory Review, 24(3), 307–326.
https://doi.org/10.1080/13264826.2021.1925716
Das, A., Brunsgaard, C., & Madsen, C. B. (2022). Understanding the AR-VR Based
Architectural Design Workflow among Selected Danish Architecture Practices.
Proceedings of the International Conference on Education and Research in Computer
Aided Architectural Design in Europe, 1, 381–388.
https://doi.org/10.52842/CONF.ECAADE.2022.1.381
Fologram. (2021). https://fologram.com/
Hale, J. A. (2017). Merleau-Ponty for architects. Routledge, Taylor & Francis Group.
Hermund, A., Klint, L., & Bundgård, T. S. (2018). The Perception of Architectural Space in
Reality, in Virtual Reality, and through Plan and Section Drawings: A case study of the
perception of architectural atmosphere. In A. Kepczynska-Walczak, & S. Bialkowski
(Eds.), Computing for a better tomorrow: eCAADe 2018 (Vol. 36, pp. 735-744).
http://papers.cumincad.org/cgi-bin/works/paper/ecaade2018_141
Hostetter, A. B., & Alibali, M. W. (2008). Visible embodiment: Gestures as simulated action.
Psychonomic Bulletin & Review 2008 15:3, 15(3), 495–514.
https://doi.org/10.3758/PBR.15.3.495
Milgram, P., & Kishino, F. (1994). A Taxonomy of Mixed Reality Visual Displays. IEICE
Transactions on Information and Systems, E77-D(12), 1321–1329.
Mindrup, M., & Wells, M. (2020). The Architectural Model as Tool, Medium and Agent of
Change. Architectural Theory Review, 24(3), 221–223.
https://doi.org/10.1080/13264826.2020.1918914
Pallasmaa, J. (2012). The eyes of the skin: architecture and the senses. Wiley-Academy ; John
Wiley & Sons.
Prithul, A., Adhanom, I. B., & Folmer, E. (2021). Teleportation in Virtual Reality; A Mini-
Review. Frontiers in Virtual Reality, 2, 138.
https://doi.org/10.3389/FRVIR.2021.730792/BIBTEX
RIBA, & Microsoft. (2018). Digital Transformation in Architecture. NBS Research, 1–46.
https://www.architecture.com/-/media/gathercontent/digital-transformation-in-
architecture/additional-
documents/microsoftribadigitaltransformationreportfinal180629pdf.pdf
Rowe, P. G., & Bechthold, M. (2017). Design thinking in the digital age. In The incidents.
Sternberg Press.
Sun, L., Fukuda, T., Tokuhara, T., & Yabuki, N. (2013). Difference between a physical model
and a virtual environment as regards perception of scale.
DOI:10.52842/conf.caadria.2013.457