Rethinking Smart Home Design: Integrating Architectural Perspectives and Technologically-driven Design Thinking within a Framework Archi Dasgupta Dissertation submitted to the Faculty of the Virginia Polytechnic Institute and State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Computer Science and Applications Denis Gračanin, Chair Douglas A. Bowman R. Benjamin Knapp James R. Jones Krešimir Matković August 9, 2021 Blacksburg, Virginia Keywords: Smart Built Environment (SBE), Smart Home, Technology, Architecture, Design, Internet of Things (IoT), Human Computer Interaction (HCI), Human Centered Design (HCD), Ambient Intelligent Environment, Human Building Interaction (HBI) Copyright 2021, Archi Dasgupta
204
Embed
Rethinking Smart Home Design: Integrating Architectural ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Rethinking Smart Home Design: Integrating ArchitecturalPerspectives and Technologically-driven Design Thinking within a
Framework
Archi Dasgupta
Dissertation submitted to the Faculty of the
Virginia Polytechnic Institute and State University
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
in
Computer Science and Applications
Denis Gračanin, Chair
Douglas A. Bowman
R. Benjamin Knapp
James R. Jones
Krešimir Matković
August 9, 2021
Blacksburg, Virginia
Keywords: Smart Built Environment (SBE), Smart Home, Technology, Architecture,
Design, Internet of Things (IoT), Human Computer Interaction (HCI), Human Centered
Design (HCD), Ambient Intelligent Environment, Human Building Interaction (HBI)
Copyright 2021, Archi Dasgupta
Rethinking Smart Home Design: Integrating ArchitecturalPerspectives and Technologically-driven Design Thinking within a
FrameworkArchi Dasgupta
(ABSTRACT)
Smart homes, equipped with sensing, actuation, communication, and computation capabili-
ties, enable automation and adaptation according to the occupants’ needs. These capabilities
work together to build holistic spatial and living experiences for the occupants. Smart tech-
nologies significantly impact spatial experiences, making smart home design an architectural
problem along with a technological problem. Nevertheless, smart home research focuses pri-
marily on standalone technological solutions, where the spatial/architectural aspect is largely
absent. We argue that addressing the technological aspects isolated from the spatial context
leads to reduced experiences for the users/occupants, as this practice blocks the pathways
to develop holistic and innovative smart home solutions. Hence, we focus on bridging the
gap between architectural and technological components in smart home research. To this
end, we studied the design of smart homes from related disciplines, i.e., architecture, human-
computer interaction, human–building interaction, industrial manufacturing, and modular
assembly. Our research used the triangulation technique to consult with subject matter ex-
perts (researchers, practitioners, and professors of related disciplines) to understand current
design processes. We conducted ethnographic studies, focus group studies, and in-depth
interviews and identified challenges and best practices for smart home design process. Our
investigation recognizes a nascent research problem where the technological and architec-
tural aspects come together in the design thinking of smart home designers. We expanded
the scope of design thinking to include three primary elements of smart homes- embed-
ded technology, architectural elements, and occupants’ needs. This multidisciplinary and
complex process requires a well-defined design framework to methodically address all the
issues associated with it. Hence, we developed a user-centered design framework, ArTSE,
through an iterative Delphi study to guide the smart home design process. ArTSE stands
for “Architecture and Technology in Smart Home DEsign”. This framework guides user
requirements collection using HCI models, technology decision making, interaction modal-
ities selection, the decision support system for schematic design, technology infrastructure
development, and production of the necessary documentation. This framework is an evolu-
tion of the normative theory in the architectural design process that caters to the needs of
smart home design. For defining implementation strategies, we applied the framework to a
case study– a smart reconfigurable space design project. Overall, we document different as-
pects of the smart home design process and provide a comprehensive guideline for designers,
researchers, and practitioners in this area.
Rethinking Smart Home Design: Integrating ArchitecturalPerspectives and Technologically-driven Design Thinking within a
FrameworkArchi Dasgupta
(GENERAL AUDIENCE ABSTRACT)
Smart homes have automation systems so that occupants can monitor and control lighting,
heating, electronic devices, etc. remotely using phones/computers. Smart home devices
and components are equipped with sensing, actuation, communication, and computation
capabilities, to enable automation and adaptation according to the occupants’ needs. These
capabilities work together to build holistic spatial and living experiences for the occupants.
Smart technologies significantly impact spatial experiences, making smart home design an
architectural problem along with a technological problem. Nevertheless, smart home research
focuses primarily on standalone technological solutions, where the spatial/architectural as-
pect is largely absent. We argue that addressing the technological aspects isolated from
the spatial context leads to reduced experiences for the occupants, as this practice blocks
the pathways to develop innovative smart home solutions. Hence, we focus on bridging
the gap between architectural and technological components in smart home research. To
this end, we studied the design of smart homes from related disciplines, i.e., architecture,
human-computer interaction, human–building interaction, industrial manufacturing, and
modular construction. We consulted with subject matter experts (researchers, practitioners,
and professors of related disciplines) to understand current design processes. We conducted
ethnographic studies, focus group studies, and in-depth interviews and identified challenges
and best practices for smart home design process. Our investigation recognizes a nascent
research problem where the technological and architectural aspects come together in the de-
sign thinking of smart home designers. We expanded the scope of design thinking to include
three primary elements of smart homes- embedded technology, architectural elements, and
occupants’ needs. This multidisciplinary and complex process requires a well-defined design
framework to methodically address all the issues associated with it. Hence, we developed a
user-centered design framework, ArTSE, through an iterative procedure to guide the smart
home design process. ArTSE stands for “Architecture and Technology in Smart Home
DEsign”. This framework guides user requirements collection using HCI models, technology
decision making, interaction modalities selection, the decision support system for schematic
design, technology infrastructure development, and production of the necessary documen-
tation. For defining implementation strategies, we applied the framework to a case study–
a smart reconfigurable space design project. Overall, we document different aspects of the
smart home design process and provide a comprehensive guideline for designers, researchers,
and practitioners in this area.
v
Dedication
To my parents (Dasgupta Asim Kumar and Sumana Gupta),
who gave me wings to fly.
To my siblings (Urmee, Shamit), dearest friends, and my advisor, who were the wind
beneath my wings.
vi
Acknowledgments
I am forever grateful to my parents for being my biggest supporters. They inspired me to
live, love, and laugh through even the toughest of times. They gave me the courage to dream
and the confidence to pursue those dreams against all odds. My heartfelt gratitude to my
siblings, my partners in crime, for always lifting me up and giving me strength. This journey
would not have been possible without the support and encouragement of my adoring family.
I would like to thank my advisor, Dr. Denis Gračanin, for believing in me. I took a big
leap of faith by changing the course of my academic path, he was the one who guided me
through the ups and downs with extraordinary patience. My sincerest thanks to my com-
mittee members who have always encouraged me and helped pave the pathway.
It was a joyful ride from the beginning to the end, thanks to my beloved friends. I would like
to acknowledge the constant support from my dearest friends (Sabrina Afrin, Bushra Taw-
Architecture 10 4 PhD Student Digital Fabrication Design,Material and Assembly
Table 4.1: Participants’ profiles for Delphi studies (reproduced from Table 3.1).
After this, the moderator describes the proposed design framework to the participants.
4. Please briefly discuss your opinion and suggestions on the proposed framework (Fig-
ure 4.3, Figure 4.6).
As a baseline for the framework, we build on our prior work (Figure 4.2 (Right)) [50] and
our insights from the studies described in Chapter 3. In a smart home, the underlying
technology framework enables the design of a context-aware physical environment [51, 153].
The physical environment design and traditional architectural concepts can facilitate design
thinking for defining the smart home design process. Interaction design for interfacing with
smart objects is another critical issue in smart home design [76]. To address all these aspects
together, this baseline describes a holistic, user-centric design framework for smart home
design [48, 50]. This is the most detailed design framework that addresses the architectural
elements, technology aspects, and user’s perspectives for smart environment design to the
best of our knowledge [50, 90].
Baseline Framework — The traditional architectural design process is depicted in Fig-
ure 4.2 (Left) and the baseline framework in Figure 4.2 (Right) [50]. The baseline framework
76 Chapter 4. Iterative Development of a Smart Home Design Framework
Figure 4.2: Left: Traditional architectural design process. Right: Baseline frameworkfor smart home design. We adopted a color code scheme for different phases where Yel-low represents Schematic Design, Blue represents Design Development, Orange representsPresentation & Evaluation, and Green represents Construction (reproduced from [50]).
divides the design process in 4 phases —
1. Schematic Design – Determining the basic scheme of the project based one user re-
quirements.
2. Design Development – Detail development of the design along with technology inte-
gration.
3. Presentation & Evaluation – Presenting the design to clients and stakeholders and
finalizing the design through an iterative process.
4. Construction – Producing working drawings and construction of the design.
We adopted a color code scheme for different phases where Yellow represents Schematic
Design, Blue represents Design Development, Orange represents Presentation & Evaluation,
and Green represents Construction (reproduced from [50]). We followed a similar scheme
throughout the document. Each of the phases are divided into steps that guide through
4.1. Developing First Iteration of the Proposed Framework 77
the smart home design process by integrating both architectural and technological concerns.
The primary difference between the traditional architectural design process and the baseline
framework is the inclusion of steps for technological concerns.
In phase 1 (schematic design), step 1.3, the baseline framework suggests using HCI models
to gather user data on technology preferences. Phase 2 (design phase) is also elaborated
further by including additional steps that are unique to smart home design —
• Step 2.2 – Technology Integration: This step consists of designing the technology
infrastructure based on the requirements defined in the first phase.
• Step 2.3 – Interaction Techniques: This step consists of designing interaction techniques
to control the smart devices and functionality.
• Step 2.4 – Data Integration: This step consists of designing the underlying data col-
lection, storage, and analysis system.
• Step 2.5 – System Architecture for Underlying Technology: This step consists of final-
izing the technology stack.
Phase 3 consists of presentation, client feedback, and evaluation. Phase 4 is for construction
which includes steps for detail development, working drawing, and construction.
4.1 Developing First Iteration of the Proposed Frame-
work
We gather feedback on the baseline framework (Figure 4.2 (Right)) during the three focus
group meetings conducted through February–April 2020 (Figure 4.1). We developed the first
78 Chapter 4. Iterative Development of a Smart Home Design Framework
iteration of our framework based on these feedbacks and by examining the process laid out
by RIBA [126], AIA [142] and literature from Lawson [107].
During the focus group discussions, participants p4 and p5 elaborated on the four phases of
design that they typically follow [107]:
1. Assimilation – Information collection about project requirements.
2. General study – Schematic design and idea generation.
3. Development – Detailed design development.
4. Communication – Conveying the design through drawings and renderings to clients
and other stakeholders.
While discussing the baseline framework, P4 suggested,
“Architecture projects are time consuming and sometimes go on for more than a
year, so it is important to be able to go back to the information collection process
from the other steps. Moreover, different levels of smartness are possible, so client
feedback is important for each step to address the specific needs of occupants.”
We conclude that the workflow of the framework should mirror the iterative nature of the
work. Participants mentioned the importance of prototype building for testing especially
in the case of smart environment. Participants discussed a challenging aspect of smart
environment design as architects,
A big hurdle for us while designing a smart home was bridging our knowledge
gap for technology design.
4.1. Developing First Iteration of the Proposed Framework 79
Figure 4.3: Iteration 1 of the proposed framework.
Smart home designers are in need of a decision support system for gathering information
about the available smart technology and choosing appropriate options considering the com-
parability between different products. Cost estimation is necessary for scoping out the
project. Participant p2 also suggested that “Phase 4” should be “Implementation” instead
of “Communication” as communication with the client is actually a continuous task through-
out the design process.
The First Iteration of the Proposed Framework— We developed the first itera-
tion (Figure 4.3) of our framework addressing the findings from the study. In the diagram,
each box represents an activity or function. The bidirectional arrows represent a two-way
relationship between the activities. The dashed arrows represent an optional relation and
the solid arrows represent a recommended relationship. We also use a “plus” (+) symbol to
mark the steps that exist in the traditional architectural process but significantly change in
SBEs, and a “star” (*) symbol to mark the steps that are unique to SBEs.
Contrary to the baseline, our workflow is iterative instead of sequential to emphasize on
80 Chapter 4. Iterative Development of a Smart Home Design Framework
Figure 4.4: Iteration 2 of the proposed framework.
Figure 4.5: Iteration 3 of the proposed framework.
4.1. Developing First Iteration of the Proposed Framework 81
the iterative nature of the work. Additionally, we emphasize more on the detailed design
activities and client feedback oriented process in phase 2 and phase 3 to better serve both
the designers and occupants. We have also structured the main phases and steps differently
to support detailed design processes. The basic idea for each phase is briefly discussed in
this section, a detailed description for each step is described in a later section along with the
final framework, ArTSE (Section 4.3).
We divided the design process into the following four phases—
1. Ideation – The first phase is dedicated to assimilating client’s requirements and other
geographic/climatic data. We suggest using HCI models to understand the client’s
activities of daily living. We also include “client feedback” as an important part of
each phase.
2. General Study – The second phase mostly consists of schematic design and making
technology decisions like which smart functionality will be provided, how does it affect
the overall spatial design, multimodal interaction techniques (voice, gesture, touch-
screen), etc.
3. Development – The third phase is where the design team develops the details of imple-
mentation. This phase consists of technology architecture and spatial infrastructure
design and testing them using prototypes.
4. Implementation – The fourth phase is dedicated to finalizing the design and drawings
and moving on to construction.
82 Chapter 4. Iterative Development of a Smart Home Design Framework
4.2 Process of Finalizing the Framework
We developed the second (Figure 4.4) and third iteration (Figure 4.5) of the framework based
on the Delphi studies. We describe the iterative development process in this section.
Developing the second and third iterations of the framework: Del-
phi Study Round 1 & 2, In-depth Interviews, Focus Group Discussions–
We modified the first iteration through two rounds of Delphi studies (Table 4.1), five in-
terviews with subject matter experts (Table 3.3), and a focus group discussion with Group
2 (Table 3.1) from May–September 2020. Throughout this process, we developed the second
and third iterations of the framework. The appendix contains the diagrams showing the
incremental development of the proposed framework (Appendix A).
Feedback and Discussion– The idea of a fully equipped smart home is gaining traction
more recently. The participants discussed that educating the occupants about available
technology options and benefits is crucial for a new concept like smart home to take off.
Participants also put much emphasis on adopting an user-centered design approach to en-
sure success. While discussing how this idea can become widely adopted, participant p12
mentioned that, in the USA more than 90% of the homes are developed by builders and they
are well positioned to offer smart homes as a service for making it widely adopted. He also
suggests,
“...this might be a more lasting effect of COVID...The shift to teleworking means
that we will see more and more automation and technology in homes.”
4.2. Process of Finalizing the Framework 83
Participants p2, p5, and p14 suggested that client feedback should be part of a continuous
feedback loop for every step in each phase. Participant p10 wondered if depending more
on the designer’s expertise and less on the client’s wishes is a better idea because they
might not always know a better solution. Participant p14 argues that with the availability
of increasingly efficient solutions like thermal enclosures and HVAC systems, the energy
efficiency of smart buildings depends more on user behavior. However, the current design
or construction practices do not follow a user-centered approach. It is a more waterfall-type
sequential approach. Participant p14 quipped,
“Current smart building construction practices are an antithesis of user-centered
design.”
The study participants validated that our work is going in the right direction and provided
suggestions for improvement. We developed the final version of the framework, ArTSE,
based on these suggestions.
Developing the final framework, ArTSE: Delphi Study Round 3–
A final round of the Delphi study with group 1 (Table 4.1) and two more interviews are
conducted from October—November 2020.
We discussed the following questions during the final round of Delphi study–
• Is it a significant contribution to the body of knowledge?
• Is it a significant advancement over what is available now?
• Was it able to capture the design process and additional requirements?
84 Chapter 4. Iterative Development of a Smart Home Design Framework
• Will you be willing to use this framework in a future smart home/environment design
project?
• Anecdotal comments.
Suggestions– We incorporated suggestions from the study participants into the final frame-
work, ArTSE. Participants p1, p10, and p11 mentioned that the maintenance/update step
needs to be considered in the last phase to emphasize on the need of sustainability of tech-
nology. Participant p1 suggested extending the framework to be one more layer deeper. This
layer (the knowledge layer) includes the tacit, explicit, and procedural knowledge about the
domain to facilitate the designer. For example, this layer discusses existing technology so-
lutions, expected input and outputs for each step, existing interaction modalities, etc. The
technological aspect is discussed in detail in Chapter 5. Participant p1 mentioned that the
inclusion of information about necessary technologies, inter-operability issues, examples, etc.,
within the framework will make it a contribution to the body of knowledge.
Participant p4 suggested developing a tool for client feedback that can assist in providing
qualitative and quantitative feedback to clients. Quantitative feedback incorporates the cost,
energy usage, etc. and qualitative feedback incorporates the information about the effect on
wellness and quality of life. Additional documentation, contractual agreements, or drawing
requirements can also be included within the steps.
Participants p1 and p4 also suggested providing a database of information to support decision
making, including links to other resources. Participant p5 commended the framework and
thanked the researchers for developing and sharing the framework. Participant p5 also
suggested showing which steps belong specifically to smart home design and which ones
were present in both smart environment and traditional architectural design.
4.3. Final Framework: Architecture and Technology in Smart Home DEsign (ArTSE) 85
Figure 4.6: Final Framework: Architecture and Technology in Smart Home DEsign(ArTSE).
4.3 Final Framework: Architecture and Technology in
Smart Home DEsign (ArTSE)
Based on the Delphi study suggestions, we elaborated the steps by including detailed design
activities (Figure 4.6). For phase 2 and phase 3, the circular layer represents two layers of
activity– the cognitive layer of making design decisions and the outer layer of communicating
with the external consultants and clients. We used the Integration Definition (IDEF) for
Function Modeling as a graphical presentation technique (Figure 4.7) [13, 135]. We discussed
the necessary inputs, outputs, controls, and mechanisms [13] while describing these activities
in detail–
• Inputs – Objects and/or data needed to perform this activity.
86 Chapter 4. Iterative Development of a Smart Home Design Framework
Figure 4.7: IDEFo’s graphical format (adapted from [13]).
• Outputs – Results or documents created by completing this activity.
• Controls – Standards, plans, templates, specifications, etc.
• Mechanisms – Necessary tools and resources to complete this activity. For example,
people with specific skillsets, specialized equipment, etc.
The computational and physical infrastructures are considered interdependent from the be-
ginning of the design process. Phases 1 and 4 are more sequential, whereas phases 2 and 3
are more iterative, consisting of two layers of activities based on the suggestions from par-
ticipant p1. Participant p13 noted that all phases do not need to be circular, for example,
we do not want to spend too much time on phase 1. Moreover, different design tasks can
be at different phases of the design process at a given time. For example, the spatial design
task can be in phase 3 at the time when the technology design task is in phase 2. For ex-
ample, at the time of the discussion, the spatial layout design task for the Project 1 was in
phase 3 (Development), whereas the technology design task was in phase 2 (General Study).
Another important suggestion from p4 is to include client agreement with client feedback so
that there is no surprise with design decisions. This is applicable for all phases.
4.3. Final Framework: Architecture and Technology in Smart Home DEsign (ArTSE) 87
Figure 4.8: The Ideation process.
4.3.1 Phase 1: Ideation
We layout the steps of phase 1 so that the design goals can be identified from the beginning
of the design process. A multidisciplinary team assembly leverages technology, spatial design
and building construction expertise. Client feedback loop at each step is crucial for an user-
centered design framework. The inputs, outputs, controls, and mechanisms for this phase
are–
• Inputs – Clients, program requirements, budget, etc.
• Output – Developing design requirements, concept, and cost estimates.
• Controls – Standards, plans, HCI models, etc.
• Mechanisms – Assembling a team with architects, technology consultants, computer
scientists, and engineers.
1.1 – Initial Program Analysis. The program requirements, budget, timeline and addi-
tional data– like location data, climate data, site information, etc. needs to be gathered in
88 Chapter 4. Iterative Development of a Smart Home Design Framework
Figure 4.9: Site analysis using 2D graphics (taken from [47, 48]).
the beginning (Figure 4.9). For example– to tackle the heat from dessert climate, Project 2
included a shaded porch area and vernacular vegetation in the design.
1.2 – Team Assembly. Project 1 team mentioned repeatedly that the main challenge
for them was to make the technology decisions, learn about available options, compare
between different options, etc. So team assembly with subject matter experts and seeking
advice from external consultants is an important step in smart environment design process
for ensuring success of the project. The project team can be built with architects, domain
92 Chapter 4. Iterative Development of a Smart Home Design Framework
Figure 4.12: The General Study process.
4.3. Final Framework: Architecture and Technology in Smart Home DEsign (ArTSE) 93
Smart Functionalities Benefits
Controlling HVAC Energy, Comfort, Health
Controlling Security Appliances Safety, Health
Controlling Comfort Comfort, Health
Table 4.2: Capabilities and benefits of smart functionalities.
2.1 – Scheme Design. Output from the first phase is used to develop a design scheme
(Figure 4.13) based on the functional requirements and the primary focus area (e.g., energy
conservation, healthcare, etc.). The design team determines the primary focus area based on
the client’s requirements. This decision influences the overall technology choices, interaction
design, and spatial design. Table 4.2 shows how smart functionalities impact the quality of
life. Available smart home technologies can be divided into the following broad categories,
e.g.,–
• Lighting – Remotely controlling lights using voice commands, gestures, or mobile apps.
• Security/Safety – Surveillance systems, occupancy detection, wearable technology, etc.
• Thermal comfort – Remotely controlling HVAC systems using different interaction
modalities.
• Convenience – Height-adjustable fixtures, voice-assistants, assistive robots etc.
The design team can help clients make an informed decision by providing a comprehensive
overview of how each category will make an impact (Table 4.3) on the way of life. For
example, the space might respond to the user’s presence by automatically turning the light
on/off; new LED lighting systems can assist in regulating the circadian rhythm; the quality
of space can be enhanced by integrating a certain category of technology, etc.
94 Chapter 4. Iterative Development of a Smart Home Design Framework
Figure 4.13: Schematic design example. Left: Activity based layout. Right: Smart technol-ogy inclusion with spatial layout
Metrics defining quality of lifeQuantitative Qualitative
Cost Effect on activities of daily living
Maintenance/update efforts Effect on health & wellness
Operation Quality of space
Energy Use User experience
Table 4.3: Metrics for measuring the impact of each category of technology on the qualityof life– a decision support system for both the clients/occupants and design team.
4.3. Final Framework: Architecture and Technology in Smart Home DEsign (ArTSE) 95
The design team needs to get a green signal from the client before going into detailed
design. An overall spatial design (schematic floor plan), smart technology scheme, and
massing scheme based on activity flow diagrams and area requirements are drafted in this
step (Figure 4.13).
2.2 – Technology Decision. The broad primary focus (e.g., convenience, health technol-
ogy, security/privacy or energy efficiency, etc.) is a determining factor for making technology
decisions. We provided a comprehensive overview of the underlying technology previously
in subsection 2.2.2.
Augusto et al. defined a smart home or ambient intelligent environment as “...a digital en-
vironment that proactively but sensibly supports people in their daily lives” [24]. Smart
spaces gather data about the state of smart objects using sensors and respond to changing
conditions and user-interaction. Interconnected communication between every day objects
is necessary to facilitate such environments, which can be achieved by IoT. Therefore, the
technology system design requires successfully combining the heterogeneous sensors, actua-
tors, and devices with a software platform to develop a responsive environment and smooth
user experience.
Participants p4 and p5 mentioned that, after deciding on the technology and approach,
technology candidate choice was one of the biggest hurdles during their design process.
They explored questions like how to choose which technology, where to get appropriate
information, how to compare between available options, and what are the new expertise
needed to be taught to smart home designers. They developed and used a “choosing by
advantage” technique for choosing vendors/providers [64]. We build on this work and define
qualitative and quantitative metrics as a decision support system (Table 4.4). The metrics
can be assigned different weights based on the preference of the clients and the design team,
and the weights may change per project. For comparing the vendors/providers, the metrics
96 Chapter 4. Iterative Development of a Smart Home Design Framework
MetricsQuantitative Qualitative
Price Reliability
Capabilities UI & User Experience
Compatibility Ease of Operation
Market Penetration Cultural Adaptation
Energy Use Ease ofMaintenance/update
Table 4.4: Criteria for choosing smart technology vendor/provider– a decision support systemfor both the clients/occupants and design team.
are quantified and a final score is obtained from their weighted sum. Criteria definitions for
the decision support system are as follows–
• Price – Cost of main equipment and installation.
• Capabilities – All functionalities and capabilities, energy efficiency, etc.
• Compatibility – Ability to use the product alongside other products.
• UI & User Experience – User friendliness of the UIs.
• Cultural adaptation – Whether the product raises any cultural concern.
• Market penetration – Whether the product is readily available in the target market.
• Reliability – Whether the manufacturer can be relied on to be operational for at least
the next decade.
These metrics can be useful for suggesting technology/vendors based on the client’s criteria
and categorizing technology/vendors. If any vendor provides good enough functionality at
4.3. Final Framework: Architecture and Technology in Smart Home DEsign (ArTSE) 97
Figure 4.14: From left to right: (a) Gesture-based UI using Kinect to control smart lights,(b) MR-based UI, user’s POV (c) Voice command based UI, (d) GUI (OSRAM Lightifyapp). [76, 77]
a better cost, they can be chosen. For example, these comparison metrics can be used to
choose a lighting vendor from Philips Hue, Ring by Amazon, C by GE, Nanoleaf light panels,
etc.
For information collection, verification, and comparison, designers can use market research
findings, social trends, sites consisting of consumer reports, reports from sources like Stanford
Urban Informatics Lab (UIL), Chicago, and corresponding magazines in a particular area
(lighting, HVAC), etc.
2.3 – Interaction Design. There are multiple modalities for user interaction with SBEs.
Interaction modalities can be device-based (switches, GUI, input devices, etc.), where the
user monitors and controls the smart environment through a UI. On the other hand, in-
teraction can be done by utilizing the full capabilities of the human body (gesture, voice
commands, etc.), where the smart environment reacts to device-free spontaneous user ac-
tions [71, 76, 103].
Participants p8 and p13 suggested that making the interaction simple, intuitive, and acces-
sible are the primary challenges. For example– having to navigate through too many pages
in an UI for performing a simple task might frustrate users. Interfacing with a smart home
needs to be straightforward to put less cognitive burden on users [90]. Exploring an overall
98 Chapter 4. Iterative Development of a Smart Home Design Framework
acceptable level of intrusion from voice assistants/automated systems is an important part
of the design process.
We have explored how to design interactive and engaging user experiences with digital sys-
tems and SBEs through the design and implementation of interaction techniques that lever-
age multimodal embodied interactions [52, 76]. To understand the user interaction with
SBEs, we conducted a comparative study comparing four interaction modalities. Figure 4.14
depicts a comparison of four interaction modalities, i.e., Voice-based, MR-based, smart-
phone GUI-based and gesture-based interface, to compare their learnability, efficiency, and
memorability. Different interaction techniques were deemed suitable for different tasks de-
pending on the complexity and context. Our analyses suggest that a multimodal approach
is preferable to a uni-modal approach as it can leverage different techniques for different
contexts [76, 77]. Since the novel interfaces were as well received as the existing interfaces,
we suggest that future research should further explore various novel interaction techniques
to develop efficient multimodal approaches. We provide a more detailed discussion in Sec-
tion 5.2.
2.4 – Cost & Time Estimate. Budget is arguably the single most influential factor that
shapes the architectural and technology aspects of smart home design. Especially, in the
case of SBEs, there are many levels of smartness available with varying degrees of expenses.
One possible approach could be to provide an incremental standard, advanced and premium
level of smartness. The higher-end scenarios will have additional functionality, for example,
pricier Samsung fridges have monitors. A fourth category can also be conceptualized to
support elderly and differently abled people (ageing-in-place and home health care).
Even though there are multiple dimensions to technology selection, an useful functionality for
the clients would be to be able to choose different technology/ vendors based on the estimated
cost. This would help the clients to make an informed decision. Our conceptualized tool
4.3. Final Framework: Architecture and Technology in Smart Home DEsign (ArTSE) 99
Figure 4.15: Concept diagram of a tool for clients/occupants for vendor selection throughcost analysis.
can facilitate clients/occupants to choose products/vendors through drag and drop methods
and see the estimated cost (Figure 4.15). This can also be further extended to include other
metrics.
Phase 2 marks the design freeze for the project. It is crucial to evaluate the cost and timeline
after the schematic design and to have a client agreement. This will help avoid surprise or
denial and reduce the chances of having to go back to the design board.
4.3.3 Phase 3: Development
This phase consists of developing detailed designs based on the outcomes of phase 2. Tech-
nology decisions can affect the physical design. Some of the steps in this phase can be broken
down further to its own iterative processes. The inputs, outputs, controls, and mechanisms
for this phase are–
• Inputs – Schematic design, interaction and technology schemes, etc.
5.1. A Reference Implementation of Technology Infrastructure [153, 155] 115
periodic analysis based on machine learning algorithms to process the stored data. Our
framework allows easy integration of analytical methods into the system to generate models
capable of predicting user behavior based on time series data. We provided a web interface for
the occupants to visualize the current status of devices and the results of the analyses. Time
series data can be viewed as graphs for previous hours, months, etc. The interface includes
a curated PI Vision dashboard to visualize the state of the smart devices and readings from
the sensors in real time. The interface also provides the capability of remotely controlling
the devices/appliances of the smart space. The occupants can use the interface to remotely
control the devices as a response to the visualization.
5.1.3 Case Studies
Providing support for AmI environment requires a stable infrastructure that provides sen-
sitivity and responsiveness, and is context-aware to provide intelligence and adeptness. We
described three case studies to explore the various functionalities of the proposed infras-
tructure. The purposes of these case studies are to check the stability of the system over
a long duration of time and explore machine learning approaches towards creating an AmI
environment.
Case Study 1: Exploring System Stability
A dedicated room used for case studies and experiments was prepared to continuously gener-
ate heterogeneous data over the period of 20 days for the first set of sensory devices and five
days for the second set. We divided the sensory devices into two categories: environment
monitoring units and energy consumption units
The sensory units used for monitoring the study space are as follows:
116 Chapter 5. Technological Aspects of the ArTSE Framework
Processing Unit
MQTTPublisher
MQTTBroker
Processing Unit UFL connectorfor historical data
Restful Web Application
Extract Using AF SDK
Real-time Analysis
PI System
MQTTSubscriber
MQTT to OMF
Storin
g analysis resu
lts
Photo resistor
Barometer
360 Lidar
Unidirectional Lidar
Thermistor
Current Monitoring
Data Collection Data Storage
Data Analysis
Data Collection setup
Figure 5.1: Study setup.
• One barometer for measuring the pressure, temperature, and altitude of the room
@1Hz
• One thermistor to measure room temperature with higher accuracy @2Hz
• One photo resistor to measure the intensity of the light inside the room @2Hz
• One Unidirectional Lidar for detecting a person entering the room @10Hz
As an energy consumption unit, a current monitoring controller was used to monitor four
devices, two server desktops, and two monitors, over a period of five days.
In this implementation, we are using two Raspberry Pis, as our processing unit to convert het-
erogeneous sensory data into homogeneous data using MQTT protocol. Another Raspberry
Pi is used as a MQTT broker. Figure 5.1 shows a diagram of the implemented setup.
The PI server subscribes to the topics, converts MQTT to OMF messages, and stores the
data into the database. There are two options to monitor a data. The user interface uses the
5.1. A Reference Implementation of Technology Infrastructure [153, 155] 117
RESTful approach to provide a web-based user interface. Figure 5.2 (Left) shows the graph
depicting real-time temperature, light, and energy consumption data. From the graphs, it
is easy to understand the usage pattern and the correlation between different parameters.
We developed a web-based user interface that includes a customized dashboard for each
room. Occupants are able to observe periodic updates in time series data for monitoring
the smart space. Based on the observation, they can click on buttons that publish MQTT
messages to the broker, the smart devices subscribe to the relevant topics and receive the
commands. Thus, the state of the smart devices can be changed using this approach (Figure
5.2 (Right)). Users can gain access to the web interface by using any browser on the local
network using the network address.
During the 20 days of recording the data, no failure in the system was detected. No packet
loss was detected, as we used QoS 1 and wired local networking. Although, some latency
between receiving the message might have occurred. The process of retrieving 1 million
points from the database requires around 9 seconds of processing time (µ = 9.33, σ = 0.4).
The size of the data stored in the database is around 60 MB per day (µ = 60.5, σ = 3.5).
Case Study 2: Device Recognition
For implementing an AmI environment, we aim to accomplish two main tasks. First, the
environment needs to be able to recognize the different devices that are available inside the
room. Second, the environment needs to learn the patterns that might occur in different
devices and be able to predict and broadcast them. In this case study, we explored how the
system can recognize the devices.
To test our model, we fed “ACS-F2” dataset [139] into the system. The database provides
consumption device signatures over the duration of one hour. The data contains 255 home
118 Chapter 5. Technological Aspects of the ArTSE Framework
Figure 5.2: Left: User interface showing time-series data depicting temperature, light, en-ergy consumption. Right: Web interface with MQTT publisher for controlling differentdevices based on real-time data.
Figure 5.3: Left: The Confusion Matrix generated by using seven minutes of the simulatedenergy consumption signatures. Right: The Confusion Matrix generated using 15 minutesof the data.
appliances divided into 15 categories. The model retrieves the data from the simulated ap-
pliance. By using simple k-nearest neighbor classification, the system is capable of detecting
the category of the device. The results from our study show that as little as seven minutes of
appliance consumption can be sufficient to determine the appliance category with acceptable
accuracy [154]. Figure 5.3 depicts the confusion matrix for seven minutes and 15 minutes
5.1. A Reference Implementation of Technology Infrastructure [153, 155] 119
Figure 5.4: The predicted value in blue compared to real value in orange. Top Left and TopRight: Examples of true positive. Bottom Left: An example of false negative. BottomRight: An example of false positive.
of appliance energy consumption generated by the simulation. As the result shows, more
energy consumption data will enable the model to recognize the device with higher accuracy.
Case Study 3: Predicting Energy Consumption
The goal of this case study was to provide a periodic analytic system that broadcasts the
results. A periodic forecasting method was implemented to analyze the energy consumption
of solar panels. We monitored the solar panels over a period of 19 days. During this time, a
forecasting program was scheduled at 4 PM to forecast the next 14 hours, using the stored
time series data of a battery voltage as a training set. After processing, the forecast model
uses Simple Mail Transfer Protocol (SMTP) to notify users, at what time the value becomes
less than 52 Volts.
15 predictions were reported during this time. We compared the predicted values with the
120 Chapter 5. Technological Aspects of the ArTSE Framework
actual values. If the reported time was in the range of the real value within a delta time of
30 minutes, we considered the report as true positive. We excluded the first two days from
our analysis, due to the insufficient training data. Overall, nine true positives, one false
positive, and three false negatives were reported during this study. Figure 5.4 shows some
of the results comparing predictions to the real values.
5.1.4 Discussion
In this section, we describe a multipurpose, flexible IoT-based technology infrastructure for
SBE monitoring and control systems in support of AmI. The infrastructure provides a simple
way to implement different machine learning applications that can be used to analyze the
stored data. Three case studies were used to explore the stability and potential of machine
learning approaches using the SBE sensor data. The results of the first case study show that
the system is stable and it can collect data over a long period of time without failure. The
process of retrieving one million data points of stored data requires around nine seconds.
This makes the infrastructure capable of providing a periodic analysis on the data that can
be used for training purposes.
The infrastructure provides an intuitive way to add multiple analytical methods to process
the recorded data. In the second case study, the system was capable of predicting the
category of simulated appliances based on their energy consumption signature using “ACS-
F2” database [139]. This approach can be used inside the system to recognize different
sensory units inside the building.
Finally, in our third case study, we explored how a forecasting method can be used to predict
the outcomes based on the recorded data. The proposed infrastructure supports features
such as sensitivity, adaptability, and intelligence that are required for AmI environment for
5.2. Interaction Design: A Discussion on Four Interaction Modalities [76] 121
SBEs. Our reference implementation can be an example for developing smart technological
solutions.
5.2 Interaction Design: A Discussion on Four Interac-
tion Modalities [76]
This section explores the interaction design options in the context of our proposed framework,
ArTSE, as discussed in subsection 4.3.2– Interaction Design. With the increase in the number
of connected devices in SBEs, the level of complexity involved in interacting with these
environments increases significantly [77]. Traditional HCI techniques are not always well-
suited for SBEs and this poses some unique usability challenges. To facilitate interactions
within such technology-rich SBEs, new models and interaction interfaces need to be explored.
In a previous research, we proposed a multimodal approach for interacting with smart en-
vironments [76]. We also conducted a user study to compare the learnability, efficiency,
and memorability of four interfaces: voice-based interfaces, GUI-based, gesture-based, and
MR-based interface. Our user study experiment involved four light control tasks that sub-
jects were asked to complete using four interaction interfaces. Study subjects found different
interaction techniques to be more suitable for different tasks based on the type, complex-
ity, and context of the task. We explored the usability, learnability, and memorability of
each modality, to identify both their scope and their limitations. Learnability was tested by
observing the initial performance of users while interacting with the four UIs for the first
time. Memorability was tested by evaluating subject task performance between two study
sessions. And finally, usability was measured through a combination of qualitative feedback
analysis and evaluation of task completion time. Our analysis of the study results and sub-
122 Chapter 5. Technological Aspects of the ArTSE Framework
ject feedback suggested that a multimodal approach is preferable to a unimodal approach
for interacting with SBEs.
In this section, we discuss common and novel interaction techniques, their strengths, and
weaknesses. We suggest that novel interaction techniques need to be further explored to
develop efficient multimodal approaches along with the widely used techniques [76, 77].
Discussion
The burgeoning number of embedded smart devices poses a challenge to interaction de-
sign [110]. An SBE is capable of understanding user input through touch, voice, gesture,
thoughts, etc. An SBE is also able to provide output using graphical, audio, or MR user
interfaces. Interaction modalities can be device-based (switches, input devices, etc.), where
the user monitors and controls the smart environment through a UI. On the other hand,
interaction can be done by utilizing the capabilities of the human body (gesture, voice com-
mands, etc.), where the smart environment reacts to device-free spontaneous user actions.
Commonly used interaction techniques were developed in the world of desktop computers.
Therefore, they do not leverage the full capabilities of smart environments or the human
body.
SBEs can gather and apply contextual information in aiding users with autonomous ac-
tion [138]. However, autonomous action may prove to be inefficient and over-patronizing for
users. Users require a simple and convenient user interface (UI) for conducting their day-
to-day activities in a smart environment [103, 140]. Home environment interfaces can be
either simple distributed interfaces or can comprise of more complex interfaces [105]. Light
switches are an example of simple interfaces while TV and A/V controllers are examples
of complex interfaces. These diverse interaction scenarios in SBEs are more intricate and
5.2. Interaction Design: A Discussion on Four Interaction Modalities [76] 123
complicated because of the sheer volume of functionality and interaction opportunities that
they provide, thereby demanding that additional research be conducted in this area [40, 103].
Nowadays, Graphical User Interfaces, leveraging the ubiquity of smartphones are dominant
in supporting user interaction with smart devices [103, 140]. GUIs provide a readily available
user interface as smartphones have become a part and parcel of our daily lives. GUI is more
useful for relatively complex tasks and for remotely controlling devices when the user is
not in the same physical space. However, complicated UI design can significantly increase
the task completion time even for an widely used interaction technique like GUI. Increasing
number of smart things makes it difficult for users to maintain a mental mapping of things
to apps. Having to switch between apps for different devices complicates user experience
and increases cognitive workload.
Mapping a 3D physical space to a two-dimensional (2D) layout displayed on a smartphone
screen can be tricky and may confuse users. For instance, turning a light switch on/off in
a home environment using a smartphone GUI might be seen as excessive and impractical
compared to simply using a physical light switch. One interesting functionality for future
researchers to develop would be to point the smartphone towards a smart device resulting in
the relevant app page opening up in the GUI. Including a layout plan of the built environment
within the GUI and placing device icons in the corresponding locations could also be helpful
for mapping the UI to the physical device.
Voice-based UI is gaining popularity in recent times, especially for smart home scenarios
because it resembles natural human communication. Voice-based interaction is intuitive, es-
pecially for simple tasks like controlling lights, air conditioning, etc. However, more complex
systems with various parameters pose a problem for voice-based interaction (e.g., controlling
the color and brightness of tens of light sources). Sometimes the verbal commands to inter-
act with a smart object can be too long-winded, causing users to forget these commands.
124 Chapter 5. Technological Aspects of the ArTSE Framework
Whereas, a GUI or hand gesture-based interface could prove to be faster and much simpler
for that task. In such a scenario, users would prefer using other UIs. Voice-based UI also
needs to accommodate the issues faced by non-native speakers. For example, making the
commands simpler and shorter, allowing for prolonged pauses or filler words.
Similarly, although voice-based UIs can provide a more natural way of interaction, mapping
smart devices to a set of pronounced names may not scale well with the rapidly increasing
number of devices in a smart environment. Memorizing voice commands and device names
can also introduce a considerable mental workload. Current practices of interaction design
in SBEs do not leverage the full capabilities of the human body. There is, therefore, a need
for more intuitive, seamless, and efficient interaction interfaces for SBEs [103].
The gesture-based UI is a hands-free option which frees the user from having to carry a
controller. Our study subjects were intrigued by the intuitiveness of this interface. However,
along the same lines with the findings of Kuhnel et al. [105], we conclude that gesture is more
suitable for straightforward and common interactions that have intuitive gestural perceptions
among users. For example, physically inspired gestures like “Up” and “Down” for “On” and
“Off”. The success of a gesture-based system has high dependence on the intuitiveness of
the gestures and user familiarity with the rotation direction of other interfaces like light/fan
regulators or switch on/off direction. Cultural factors (e.g., writing direction) also effect the
user’s intuition. Gesture-based interaction is more suitable for scenarios where the user and
the device are both in the same physical location.
Gesture-based interaction can provide for embodied and instantaneous interaction that lever-
ages the capabilities of the human body. In doing so, gesture-based interaction can allow
users to simply point at smart objects to control them and spare them the burden of having
to remember a plethora of complicated device names. Petersen et al. [134] evaluated the po-
tential of using gestures in their user study and determined that 80% of their study subjects
5.2. Interaction Design: A Discussion on Four Interaction Modalities [76] 125
preferred to use a gestural interface over more traditional interfaces like GUIs.
MR can be a potentially useful input modality because a smart environment is likely to have
numerous, potentially undetectable smart devices and it can be quite difficult to identify
and leverage their smart capabilities to full potential through traditional control interfaces.
The enhanced capability of MR devices can assist users in detecting and interfacing with
various smart functionalities.
A 3D digital medium like MR provides a greater amount of visual and contextual infor-
mation using holograms, lending it to be better suited for interfacing with a large number
of distributed smart objects, which would be otherwise difficult to control using traditional
interfaces. The holographic projection on top of the real environment is useful for com-
plex tasks like maintenance and assembly. Virtual indicators might be useful for indicating
proper placement of parts in case of assembly [42]. Contrarily, using a heavy, head-mounted
device at home for a simple light control task might be redundant. Even though the re-
cent MR devices are fairly light, users prefer even lighter options in the case of a wearable
device for day-to-day use in a smart home context. Other SBEs like smart factory, smart
warehouse, smart industry, etc. could be more suitable for MR-based interaction. The most
frequently used interactions need their separate buttons, gestures, or commands which are
easy to memorize or placed in a focal point of the GUI.
Overall, our findings suggest that different modalities were more suitable for different types
of tasks. SBEs consist of objects that are imbued with computation and communication
capabilities. This opens up numerous novel interaction possibilities that leverage recent
technological advances, like MR and embodied interaction. Different modes of interaction
have different strengths and weaknesses based on the task type. Hence, a multimodal ap-
proach combining novel and traditional techniques provides users with more flexible and
varied interaction options.
Chapter 6
Dissemination
In this chapter, we explore the application of the ArTSE framework using a case study to de-
velop dissemination strategies. We also aim to identify potential issues with implementation
through the case study. The case study is a research project for designing a smart recon-
figurable space (SReS) for a common area in a residential hall at Virginia Tech. The aim
is to ensure that the space is empathetic/responsive to the users’ needs. We introduced the
ArTSE framework to the project team using a manuscript and PowerPoint presentations.
The design team followed the framework throughout their design process and reported the
issues that they have faced while going through the steps. They have also published a part
of their research and discussed the use of the framework [59].
We will first discuss the limitations of our case study. A limiting factor is that the study
participants were already familiar with our research, which could have potentially biased
their feedback. The study participants are researchers working in the domain of smart
reconfigurable spaces and they come from a building construction and computer science
background which meets part of our requirements. Future research directions can include
conducting case studies with architects and builders as they are the primary target audience
of the framework. Limitations of the research also include a lack of testing outside of smart
home design. Expanding the current research to include other use cases like smart offices,
schools, etc. could open up possibilities for extending the ArTSE framework to support
other types of SBEs.
126
6.1. Case Study: The SReS Project 127
The SReS project is a suitable case study for us as our framework is aimed at residential
projects and this SBE design project is focused on designing a common area for a residential
hall. The goal of this project is to improve the efficiency of indoor space utilization by
creating an optimum layout for each activity and maximizing occupancy. As the pandemic
has put in a lot of restrictions on space utilization, this research addresses a timely concern.
This project could benefit from combining architectural and technological considerations as
it is mostly concerned about space and its maximal utilization. Since this project deals
with reconfiguring and redefining the space, architecture will play a vital role in this. This
project also needs assistance from smart technologies for physically reshaping the space,
so technological concerns are also crucial. Our framework aims to bring architectural and
technological design aspects together to offer a holistic design process for SBEs. Hence, the
project team utilized our framework during their design process.
6.1 Case Study: The SReS Project
The reconfigurable space design project was used for exploring the implementation strategies
of our framework. In this section, we describe our findings from this case study. We also
include the design team’s feedback and comments about the usability of the framework. This
preliminary application helped us in developing instruments for implementing the framework.
The Corps of Cadets at Virginia Tech are the clients of this project. The project entails
designing a common area cum lounge in one of the cadet residence halls at Virginia Tech [59].
The initial idea behind the project was that the space will reconfigure itself based on the
time of the day and usage so that it can maximize the spatial distance between seating ar-
rangements to minimize the spread of infection. In the beginning, the idea was to reconfigure
almost all components of the room, e.g., the walls and furniture. Later, the project scope
128 Chapter 6. Dissemination
was narrowed down to re-configuring the room layout and the furniture themselves. The
final outcome included creation of a layout for maximum occupancy and developing various
reconfiguration strategies.
Initially, the design team intended to follow a generic HCI design approach or activity flow.
Where the first step is to get the user requirement, then design development, building a
prototype, getting user feedback on that prototype, and then finalizing the last product.
After getting introduced to our framework, they started using the framework throughout
the design process.
We held an initial one-hour meeting (September 2020) to introduce the ArTSE framework
to the team. We have provided the framework along with the detailed descriptions as a
manuscript (Section 4.3). We have also presented the framework using presentation slides.
After that, they used our framework throughout the smart reconfigurable space design pro-
cess. We conducted a 40 minute interview (January 2021) to learn about their experiences
throughout the process.
Open ended questions for discussion:
1. Please give a brief introduction of your smart environment project.
2. What were the reasons for choosing the smart environment design approach for your
project?
3. Please provide your feedback on using the SBE framework throughout the design and
decision making process for your smart environment project.
4. What were the reasons for choosing this framework?
5. Is there any other existing design framework aimed at assisting smart environment
design process?
6.1. Case Study: The SReS Project 129
6. What do you expect from such a framework?
7. Please discuss the lessons learned and best practices.
6.1.1 Qualitative Feedback
After getting introduced to the framework, the design team mentioned that the framework
provided them a structured way of looking at the design process. As this was a research
project, some of the steps mentioned in the SBE framework were not applicable for them,
e.g., implementation, detailed drawings, etc. The team mapped their design process using
the framework and discussed what was useful or if anything was missing. They mentioned
that,
“The steps of the framework perfectly align with the necessary actions (for de-
signing a smart space).”
The first part, Ideation Phase (phase 1), consisted of coming up with ideas, collecting in-
formation through multiple interviews and meetings, and determining users for the example
implementation. The cadets have a social lounge (about 380 sqft) that they use either for
study or for company meetings. These two main configurations require different space usage.
Therefore, during the Ideation phase, the design team conceptualized changing the layout of
the room automatically to accommodate whatever activity they are doing.
The team mentioned that the SBE framework helped them realize the importance of com-
municating thoroughly with the clients,
“...the “Pitch” step is actually an important step for smart environment design
projects.”
130 Chapter 6. Dissemination
The project team had to convince the clients that there will not be “too much reconfigura-
tion” happening in that space. Some of the students were excited about the novel concept.
In the General Study Phase (phase 2), the inner layer of the cognitive process consisted of
using generative design approaches [59] for creating schematic layouts of the reconfigurable
spaces. The technology decision step consists of choosing interaction techniques for informing
users about moving objects in the space. The outer layer of the feedback loop consisted of
interviews with clients and consulting their advisor. The cost and time estimate step was
not applicable for them.
For the Development Phase (phase 3), the project team is working on two options— first
one is creating layouts with existing furniture and the second one is proposing new foldable,
reconfigurable furniture. As there will not be any prototype building, the team decided to
use virtual reality for testing the usability of the reconfigurable space by conducting a user
study.
The Implementation Phase (phase 4) is not applicable for them. To quote them,
“..but at least these first three phases of the framework we did include in the
study.”
Experience and feedback— When asked if they found the framework useful, one of the
team members responded,
“I think it was very useful, in the sense that it helped us realize a lot of things
that we were missing earlier. When we started, we had a very vague idea on
how and what we should do to come up with a solution. But, as we looked at the
framework...it helped us find all those missing pieces and put them in place....and
we are still learning as we look at it. And we will probably learn more as the
6.1. Case Study: The SReS Project 131
project progresses. ”
For example, the team have not discussed the implementation techniques for automation
yet, but looking at the framework they realize that they need to figure out the most efficient
implementation too.
One other aspect of the framework that the users liked was,
“for any project in real life, the most important constraint that comes for im-
plementing such projects is the time and the cost constraint. And I think this
framework also captures that.”
When asked if they have looked for any framework at the beginning of the design process,
the users mentioned that they were mostly relying on their and their research advisor’s
experience. When asked if they knew of any such existing framework that focuses on smart
environment design, one of the users mentioned,
“...having worked in the area of built environments and smart built environments,
for a good amount of time, I have not come across any framework that suggests
the design and development of smart built environments....In our case, we mostly
follow our prior experiences with HCI, and UI/UX design, for the design process.
But, if we had this framework to start with, then probably our design process would
have been a lot better. Nevertheless, this framework was introduced to us. And
at whatever time it was introduced to us, it did help us a lot in kind of filling up
those missing pieces and those gaps in the current project.”
We discussed the currently available guidelines for traditional built environments. One of
the participants worked in the Indian real estate development industry and he mentioned
132 Chapter 6. Dissemination
that the process guidelines that they followed were mostly a set of rules and procedures—
building codes, standard dimensions, and rules of thumb related to ergonomics. There was
no framework comparable to this. The user also mentioned,
“these rules or guidelines were not human centered...there was very little human
involvement in the design process. For example, if you have to design a room of
a certain size, then you would need to follow certain standard dimensions and
building codes, and then the room will be appropriate for a certain number of
people... and just follow this rule and design the room. And that’s it.”
We also discussed about the additional aspect of smart homes, the integration of technology,
and if the framework was able to capture that. The users mentioned that as they were
not planning to implement the project, rather simulate the smart capabilities using virtual
reality, they did not duel too much on the actual technologies to use.
“But if I think about implementing, then the selection of technology and the sus-
tainability of technology becomes very important.....because that would drastically
affect the budget, maintenance, and ease of implementation of the project....it
affects a lot of factors that are listed in the framework. ”
While discussing how the smart capabilities have impacted the design process, the users
mentioned that the decision to pursue a smart design approach has fundamentally changed
the whole design process. Increased use of technology as design tools significantly increased
because of the smart design approach. The use of generative design to figure out an optimal
layout was also adopted because of the smart environment design. Convincing the clients was
also more difficult, because people are still apprehensive of the extensive use of technology
within the built environment. We discussed about the next steps when the users start
6.1. Case Study: The SReS Project 133
thinking about the implementation of the project. Taking guidance from the framework,
they discussed that for real world objects to move around, there would be a need for sensors
and actuators that are also context-aware to alert people when things are moving.
6.1.2 Quantitative Feedback Using SUS Score
The System Usability Scale (SUS) [32], created by John Brooke in 1986, is a “quick and
dirty” usability scale consisting of 10 questions for evaluating a system. The calculated SUS
score from the users’ feedback is 90 out of a possible score of 100. This score gives us an
idea about the usability of our framework. We report the users’ responses and feedback in
We conduct a survey to gather information about SBE design goals, design processes, and
best practices. The survey takes approximately 45 minutes to complete. Users receive a 10$
Amazon gift card for participating in the study. We have included the IRB-approved user
study description and questionnaire in this section.
Title: Best Practices and Guidelines for Smart Built Environment (SBE) Design
Process Focusing on Residences
Protocol No.: IRB-20-716
Users are eligible to participate if they have previously worked in a smart built environment
(SBE) project, preferably a smart residence project. Participants have two options—
1. Completing the questionnaire asynchronously.
2. Scheduling an audio/video conference call to complete the survey in an online interview
format.
178
B.1. Questionnaire 179
B.1 Questionnaire
The following questionnaire was developed using the Qualtrics Survey Software. The user
would participate in this survey using the online tool provided by Qualtrics.
/
Summary
Q1.Information Sheet: Principal Investigator: Dr. Denis GračaninIRB# and Title of Study: IRB-20-716: Best Practices and Guidelines for Smart BuiltEnvironment (SBE) Design Process Focusing on ResidencesSponsor: Center for Human Computer Interaction User Study Funding Program
You are invited to participate in a research study. You are eligible to participate if you have previously
worked in a smart built environment (SBE) project, preferably a smart residence project.
“I am a graduate student at Virginia Tech, and I am conducting this study as part ofmy PhD research.”—Archi Dasgupta ([email protected]).
Ø WHAT SHOULD I KNOW?
If you decide to participate in this study, you can opt for one of the following options—
1. You can complete the questionnaire asynchronously.
2. You can schedule an audio/video conference call to complete the survey in an online interview
format by contacting Archi Dasgupta ([email protected]). The interview will not be recorded.
The survey aims to gather information about best practices for SBE design, goals, design processes,
and best practices. The study should take approximately 45 minutes. We do not anticipate any risks
from completing this study.
You can choose whether to be in this study or not. If you volunteer to be in this study, you may withdraw
at any time without consequences of any kind. The investigator may withdraw you from this research if
circumstances arise which warrant doing so.
Ø CONFIDENTIALITY
We will do our best to protect the confidentiality of the information we gather from you, but we cannot
guarantee 100% confidentiality.
/
Any data collected during this research study will be kept confidential by the researchers. The
interviewer will take notes to transcribe the answers and code the transcripts using a pseudonym.
Transcriptions will be uploaded to a secure password-protected computer in the researcher’s office. The
researchers will maintain a list that includes a key to the code. The master key and the recordings will be
stored for 3 years after the study has been completed and then destroyed.
Ø WHO CAN I TALK TO? If you have any questions or concerns about the research, please feel free to contact Archi Dasgupta([email protected]). You are not waiving any legal claims, rights or remedies because of your participationin this research study. If you have questions regarding your rights as a research participant, contact theVirginia Tech HRPP Office at 540-231-3732 ([email protected]). Please print out a copy of this information sheet for your records.If you would like to participate in this survey, at least 18 years old and not a student of theinvestigators, click yes to begin or no to exit.
Demographic Information
Q2. Demographic Information
Respondent's experience with SBE design
Q3. How many years of experience do you have?
Q4. Please provide the number of SBE projects you have worked on for thefollowing categories.
Yes
No
Name
Email
Occupation & Affiliation
Area of Expertise
Regular Built Environment Project
Smart Built Environment (SBE) Project
Residential
/
Q5. What was your role on the most representative smart residence project?(select all that apply)
Follow up questions on the project mentioned in Q11.
Q6.What were the reasons for choosing to construct an SBE versus a regular builtenvironment? (select all that apply)
Office
Educational
Retail
Other (please specify)
Project Name
Architectural Designer
Technology Consultant
Electrical Engineer
Mechanical Engineer
Project Manager
Civil Engineer
Construction Professional
Computer Scientist
Other (please specify)
Efficient Functionality and Convenience
Energy Conservation
Cost Efficiency
Healthcare (ageing in place, addressing disability)
Solving Spatial Limitation
Comfort
Security/Safety
Other (please specify)
/
Q7. Which types of smart functionalities do clients typically want? (select all thatapply)
Q8. Which types of smart interaction techniques do clients typically want? (selectall that apply)
Q9. How did the inclusion of smart functionality affect the architectural design?(select all that apply)
Q10. During the SBE design process and selecting smart technology, what werethe main challenges? (select all that apply)
Smart Lighting
Smart Programmable Thermostat
Smart Security System
Automated Control of Doors/Windows
Healthcare Technology
Smart Meter
Reconfigurable Space
Other (please specify)
Physical Switch
Mobile Phone Application
Voice-based Interaction
Gesture-based Interaction
Mixed Reality-based Interaction
Other (please specify)
Architectural elements (wall, doors, windows etc.) embedded with sensors/actuators
Reconfigurable spaces
Smart surfaces as interfaces
Other (please specify)
Absence of a well-defined design framework for combining smart technology design withspatial design
Absence of unified technology solutions for supporting heterogeneous smart devices
Lack of established comparative metrics for choosing smart technology
/
Q11.What are the main phases/steps that you followed during the SBE design process?(select all that apply)
Q12. In which phase was the decision made to construct an SBE versus a regularbuilt environment?
Q13. How did the decision to construct an SBE affect your design process? (selectall that apply)
Limited data source for smart technology
Other (please specify)
Ideation
Schematic Design
Design Development
Implementation
Other (please specify)
Ideation
Schematic Design
Design Development
Implementation
Other (please specify)
"Ideation" and "Design" phases changed significantly
Clients needed to be educated about smart home technologies
Needed additional steps for designing technology aspects