HEEM
Examining how AR and AI can be used to enhance
the real estate user experience
HEEM is a conceptual case study where I created a real estate mobile app that assists Tenants and Brokers/Building owners with the rental process. As someone who works in a real estate-adjacent industry, I decided to design HEEM as an app that simplifies common real estate practices that both parties would encounter, such as browsing properties, viewing in-person property showings, and scheduling showings and meetings.
As a designer, I wanted to challenge myself by introducing both Augmented Reality (AR) and Artificial Intelligence (AI) to this concept and explore how both could be effectively utilized on a mobile app.
This case study is broken down into two phases:
Phase 1
Conceptualize how AR could be used with HEEM's cameras and explore how HEEM could address common real estate pain points. How could I design HEEM to make property showings and correspondances easier for both prospective tenants and real estate brokers?
Phase 2
Explore HEEM’s capabilities with AI technology and analyze how AI could hypothetically enhance HEEM, in relation to its competitors.
Phase 1
Building HEEM's foundations and Augmented Realty camera integration
Who needs HEEM and why?
My background for developing HEEM comes from my experience working at my family’s business.
My family manages multi-unit buildings and they often correspond with both real estate brokers and tenants in the apartment rental process. Having observed this process from both the renter/broker’s side and the tenant’s side, I understand that there are distinct pain points involved, such as communication of apartment dimensions and scheduling lease signings or in-house showings between all parties involved.
Research
Gathering information and identifying HEEM's objectives
User Interviews
Before designing the mockups for HEEM, I wanted to explore these technical pain points in depth to understand what prospective users might need. And to fully understand the ins and outs of the real estate industry, I interviewed a prospective tenant and a real estate broker to understand both their perspectives and how the rental process works.
I asked the two participants questions about their rental experiences, focusing on what they listed as their criteria for showing an available unit (The Broker/business side) or deciding on a property and signing a lease. (The Tenant side) I also asked them about their thoughts on AR to determine whether or not using AR could further enhance their experiences and factor in their renting decisions.
Interview Results and Defining Focus
I've noted the following key observations (shortened) from both Tenant and Broker interviews:
"I'd love to see more transparency in the rental process, like the ability to view the apartment’s dimensions and any unit fixtures I need to know about, like previous tenant damages.
I don't like 360 virtual viewings of apartments, so I'd like to see them in person. I have a lot of big furniture, so I'd like to get real-time details about flooring, wall, and ceiling measurements, to make sure everything will fit."
"I want tenants to browse our properties in person before making a decision to sign a lease with us. We've had tenants who moved in and complained about not finding enough space for their stuff or couldn't locate things like AC units and outlets. So it would be good to have something to help tenants (and us) remember apartment details."
Following the interviews, I came to the conclusion that AR would be helpful for a real estate app like HEEM and decided to pursue this feature further on the app. Using their interviews, I broke down the participants’ needs and wants into 4 plausible features :
-
Log In and Sign Up forms, separate for prospective tenants and brokers/real estate businesses.
-
Use AR to take apartment photos, measure property dimensions, and view more information about the space in an immediate setting.
-
Book an in-person showing of the apartment and contact the broker or building managers who post the listings.
-
Upload and cast AR furniture into the apartment space for tenant’s pre-visualization purposes.
While both participants emphasized very valid and notable key points about the rental process, such as easing the communication between tenants and brokers/building owners, in the interest of this project’s short time frame I decided to focus on the app's AR camera.
" I'm open and optimistic to using AR technology for the real estate experience! It'll be cool to visually plan out where everything would go in an apartment with AR." - Tenant
" I would use AR to show information of the apartment interiors to Tenants, so I wouldn't have to measure it myself." - Broker
Define
Defining and prioritzing HEEM's key components
After deciding on which HEEM feature to prioritize, I organized this in the form of a journey map.
In the interest of time, I concentrated these journeys on the Tenant's side, because the priorities of real estate brokers and building managers are significantly different from those of a prospective tenant's.
As a way to learn about the user’s thought process, I split up the features into separate journeys that followed situational scenarios that they may face. For instance, I mapped out what would happen if a user wanted to move from one room to another, or what would happen if they wanted to use the AR camera to measure the apartment dimensions.
Mapping out these hypothetical scenarios informed the direction I eventually chose to take with the wireframes.
Low-Fidelity Wireframes
Sketching and designing HEEM's layouts for prototyping
The journey roadmaps became the foundations for HEEM's low-fidelity wireframes.
I organized my objectives for these features by hand-sketching the visual wireframes on paper.
I also sketched over pictures of real-life apartment interiors to visualize what the AR screens could potentially look like.
.png)
HEEM Process 1 - Access AR camera and upload furniture
.png)
HEEM Process 2 - Instant AR scanning of Apartment interior for results
.png)
HEEM Process 3 - Uploading own images (ex. furniture) and projecting onto AR camera view
After sketching the first drafts of the wireframes, I created the low-fidelity wireframes in Figma based on the user journeys and previous research.
I was aware that the HEEM wireframes needed to work responsively for mobile screens, so I designed the wireframes on Figma and simultaneously projected these mobile wireframes onto my cellphone to determine the appropriate sizes and information hierarchies for each page.
Prototyping and Usability Testing
Testing and refining the HEEM prototype
Once I've established the wireframes, I proceeded to build the HEEM interactive prototype based on four particular scenarios. These scenarios reflect the features and objectives that I wanted to pursue in this app, such as implementing AR in the real-estate viewing process , simplifying the process of contacting brokers, and reserving bookings to view the properties in real time.
These scenarios highlighted the following tasks:
-
Logging in and signing up for the app.
-
Using the AR camera to navigate around a property and look up property information in real-time.
-
Browsing property listings and booking a tour of a listing with a broker online.
-
Using the AR camera to cast virtual furniture into the property space for pre-visualization and customization purposes.
I contacted 2 participants who were willing to test out the hi-fidelity prototype and asked them to test these 4 scenarios. This was an important stage in the process because I needed to see if users were able to navigate and understand the layout, hierarchy, and purposes of HEEM's features.
Based on my Usability testing round, I learned the following about the new prototype:
-
Both participants thought that the log in/sign up procedure and booking a tour on the app were pretty straightforward tasks because they are commonly used forms in most mobile apps.
-
However, I manually guided my users through some of the AR related tasks because they required a previous understanding of AR and 360 camera views.
-
The participants were unfamiliar with the onscreen visual cues that appeared in the AR camera, such as the bottom navigation arrows and the camera button.
Overall, the feedback from this usability testing round was very helpful because it reminded me to focus on the app's visual affordances and how they would influence user engagement.
I learned that with a visual-heavy feature like an AR camera, I needed to lean more into designing cues that aid users in the property viewing experience, without guidance.
I also worked on establishing HEEM’s visual brand for the high-fidelity mockups. I created a style guide that served as a visual reference for HEEM’s typefaces and color palettes. I wanted HEEM’s overall look to be bold, yet uncomplicated and modern.
Phase 1 Takeaways
What did I learn? What could I have done differently? What's next?
Developing the HEEM wireframes was both an interesting and tricky experience.
Prior to this case study, I had designed most of my case studies responsively for desktop, tablet, and mobile views. So not only did I learn how to conceptualize a solely mobile app with AR capabilities from scratch, but I focused on the real estate industry. I had to research more about the basics of real estate and leasing properties from both tenant and broker/business sides to understand who I am designing for and how to make it work for them.
On a technical level, I learned how to design AR capabilities in my mock wireframes. Because AR was a challenging component to add to a mobile UX prototype, I brainstormed a variety of ways AR could be used on HEEM’s camera and navigation settings.
If I had more time to explore the HEEM app, I would address the following:
1. Expand the prototype to further address the business side of real estate.
For instance, what would happen in the app if a realtor wanted to post or take down an apartment listing?
How would the realtor respond to a tenant’s reservation via HEEM’s booking feature? And how would a realtor post and update property information using the app’s AR camera mode? I believe that there is a lot that could be further explored in this realm.
2. Conduct more usability tests that explore HEEM's current prototype scenarios.
This would include scenarios such as editing reservations and using AR markers for damages, outlets, and wiring in the apartment interiors.
Overall I felt accomplished because I challenged myself to design with AR and approach something as nuanced as the rental process.
Phase 2
HEEM and Artificial Intelligence (AI)
In recent years, Artificial Intelligence (AI) is rapidly shaping digital platforms and apps function across various industries. From enhancing user experiences to automating processes, AI is becoming a crucial component for products and businesses looking to innovate and stay competitive.
My objective in Phase 2 is to examine AI’s impact in real estate and how HEEM could learn from AI to further evolve and stay competitive in the real estate sector.
Applications of AI in the Real Estate Sector
AI revolutionized real estate and brokerage platforms by automating how potential clients view properties, interact with brokers, and decide to rent or buy. Notable brokerage platforms like Zillow and Trulia use AI to streamline the following processes:
1. Personalizing Property Recommendations
2. Automating Property Listings and Descriptions
3. Constructing Virtual Tours and Image Recognition
4. Enhancing Search and Filtering Capabilities for various properties
Looking at these processes, I am prioritizing the use of Image/Video Generative AI (or Gen AI) for the HEEM platform, and how Gen AI could enhance Virtual Staging/Tours and Property Listings.
Competitive Analysis
Looking at Real Estate tools and how they compare to HEEM?
In Phase 1 of HEEM, I highlighted the AR camera as the app’s main pre-visualization feature to project uploaded furniture, interior dimensions, and any additional apartment information (ex. Wall or ceiling materials, damages, etc.).
However Generative AI (or Gen AI) tools now have the capability to take pre-visualization and virtual staging further by creating customizable and accurate interiors in seconds. Notably, Gen AI has the capability to assist with the following in real estate:
-
Instant virtual and visual staging of properties
-
Visual tagging and identification with AI capabilities
This section will analyze how Gen AI’s benefits visual staging and virtual identification and cross-compare HEEM with 2 existing AI brokerage platforms (REimagine Home and Restb AI) to address these real capabilities.
Gen AI for Virtual Staging
.png)
.png)
REimagine Home is an AI tool that uses AI specifically for instant virtual staging. The website and app take uploaded images of home interiors/exteriors and virtually recreate the space for visualization purposes.
User can also type in personalized inputs for their image results (ex. show a fully furnished or renovated room in X or Y style) and REimagine Home would regenerate the property photo accordingly.
Notable Features
.png)
1. Uses generative AI to fill empty interiors with realistic staging - Great for Property Managers staging their properties.
.png)
2. AI assisting in visualizing empty spaces - Great for previsualizing fully furnished properties into empty spaces for new Tenants.
3. Custom AI filters - Great for Tenants, Brokers/Managers, and even property Contractors to filter and select specific property areas and styles for more previsualization and redesign ideas.
How does it compare to HEEM?
As seen from its notable features, REimagine Home utilizes generative AI as a previsualization tool for property staging. In this way, REimagine Home features are similar to HEEM's because of its ability to aid users in customizing and visually redesigning a space before renting it out.
On the one hand, I believe that HEEM could absolutely benefit from REimagine Home's generative AI because of its versatile filters and empty/furnished room staging options. For instance, if a Broker is showing a prospective Tenant the apartment space of a current Tenant who is moving out, but has not moved out their furniture from the space yet, they could easily use AI to either empty out the apartment for a clean slate look, or fill it with the prospective Tenant's furniture for reference. So HEEM's existing AR camera could implement this AI tool in this way to make the showing/staging phase smoother for both Brokers/Property Mangers and Tenants alike.
On the other hand, my concern as the designer for HEEM is that REimagine Home's AI is unable to process images and filter inputs in real time. Going back to my Tenant interview in the research phase, the Tenant interviewed said that while there are prospective Tenants who are okay with viewing virtual tours of apartments and properties, there are other Tenants (such as himself) who wish to view and visualize whether they would rent out a space by booking and touring apartments in person. And looking into REimagine Home, users are able to customize views of furnished or unfurnished properties if the pictures were already pre-uploaded into REimagine Home's website or app.
So I feel that HEEM could answer the need for instant customization for an in-person showing by combining the AI filters with HEEM's current AR cameras to redesign spaces instantly and in real time. This would give HEEM a competitive edge against competitors like REimagine Home because users could use AI to assist them in viewing apartment information and design ideas instantly, enhancing the in-person viewing process for all users.
Virtual Tagging and Identification
.png)
Restb AI uses AI to analyze property data and interpret existing property pictures.
According to their website, they focus on using AI to enhance Property Imagery to effectively market listed homes, provide context on a specific property's appraisal rates, and other essential insights on listed properties.
It should be noted that unlike REimagine Home, Restb AI is different because it is geared towards assessing property data for appraisals, inspections, and real estate marketing.
However Restb AI's current features, such as Property Tagging and Photo-based Condition Analysis, and use of AI for said features could inspire and enhance HEEM's user experience.
Notable Features
.png)
.png)
1. AI Property Tagging - Great for both Tenants and Brokers/Property Managers to instantly extract and identify property's information furnishings at a glance.
.png)
.png)
2. AI Photo-based Condition Analysis - Great for Tenants to compare and contrast the quality of property conditions with other, similar properties.
.png)
2. AI Object Detection - Great for Tenants and Property Managers to visually pinpoint specific highlights in both furnished and unfurnished spaces.
How does it compare to HEEM?
In comparison to REimagine Home, Restb AI is not a strictly previsualization AI tool, however it's use of AI to detect property highlights and conditions could be beneficial to HEEM.
A hypothetical expansion that I had in mind for HEEM was to expand its AR cameras to visually identify unit features, such as necessary outlets, air conditioning/heating units, piping, etc. This would be a useful for not only prospective Tenants who may want to quickly locate these units in person, but also for Property managers and Brokers who are showing the space but may overlook hidden or new furnishings in their properties. In this regard, Restb AI's Property Tagging and Object Detection options are visual inspirations for how I would design HEEM's expansion and add automatic unit detection.
Additionally, I am also inspired by Restb AI's Photo-based Condition Analysis because I am intrigued by how accurately their AI was able to cross-check property photos and analyze their condition quality based on similar properties.. While this particular feature is geared more towards towards property appraisals and conditions, I believe that the AI technology used in visual condition analysis could really help Tenants make in-depth comparisons of the apartments they are browsing. In HEEM's case, I envisioned using the AR cameras to scan an apartment interior in person, then having the AI automatically generate an analysis of the scanned interior and how it compares to the Tenant's other prospective apartment choices.
Additionally this AI tool would also benefit Property Managers in particular, because whenever a new Tenant moves in or an old Tenant moves out, Property Managers need to have moving in/moving out conditions on file to determine if there are any previous damages or changes they would need to replace or charge the previous Tenants. And in HEEM's case, Property Managers could use HEEM's AR cameras to project and detect property changes, save the AR images into the app, and use AI to instantly compare the recently scanned photos with previous ones. So if HEEM borrowed Restb AI's Condition Analysis AI for cross-checking purposes, Property Managers would have an easier time keeping track of their property conditions over time.
Phase 2 Takeaways and Conclusion
What did I learn? What are my takeaways?
Following my exploration of AI, I’ve concluded that adding elements of AI tools to HEEM would make it a worthy competitor. AI’s integration into real estate apps and tools has significantly enhanced the user experience for both tenants and brokers/building managers, making property searches more efficient, personalized, and accurate. And these advancements would improve HEEM’s visual experience and accessibility in the long run.
What have I learned as a designer?
Working on Phase 2 of this case study was a challenge for me because AI is relatively new (at the time of this case study) and its technologies are still in development. As a visual artist and animator, I still have many reservations about claiming generative AI as the next end-all-be-all solution to art and design. However my deep dive into real estate AI tools has introduced me to more helpful applications of generative AI, especially for apps like HEEM.
For example I found that using AI as a supplementary tool to automate the staging and previsualization stages in the rental processes could be beneficial for visual-heavy apps like HEEM. And going forward, I can see users such as Tenants and Property Realtors taking advantage of these visual AI tools to browse and make decisions on renting out properties. But while I am open to seeing more AI possibilities in UX design, I still feel that AI’s technology may miss certain things in the users’ experiences that only we, as designers and human beings, could empathize with and take into consideration. (Ex. thinking about different hypothetical situations and users that the AI may overlook)
So HEEM's phase 2 analysis taught me that UX Design is evolving with new technologies like AI and presents businesses with new ways to improve in a competitive market. However we still must advocate for ethical practices of AI when designing for the user experience. We could design for and with AI, but we will have to regulate and monitor HOW and WHY it is being used, while assessing how best to use it in our platforms accordingly.