ExplorAR Augmented Reality Creation Platform

Opportunity

Augmented Reality (AR) is becoming increasingly prevalent in a wide range of fields such as education, instructional training, entertainment, and social media. AR combines aspects of the user’s real-world environment with virtual “overlays” that display contextual information in real-time and invite user interactions, (Pietro et al., 2018). While AR has applications for industry, it also appeals to consumers who may use it to play games, try on glasses, or enhance videos on TikTok. Despite its appeal, the ability of the general population to create their own AR experiences is lagging.

Challenge

Most AR creation tools involve expensive professional software and equipment and are geared toward programmers and developers. According to Gasques et al. (2019), many AR tools “still require a high degree of programming proficiency, presenting multiple barriers and a steep learning curve for designers or non-programmers.” Leiva et al. (2020) states, “the need for programming knowledge hinders exploration and excludes non-technical designers.” We also noticed a “Catch-22” conundrum in that developers may have the programming skill to create an AR application, but they “lack the content knowledge.” (Martinez et al., 2014). Conversely, hobbyists, educators, DIYers and influencers may possess vast content knowledge but have little to no coding expertise to translate their content into an AR experience. Our team wondered, what if there was a way to have the best of both worlds?

How might we...

The motivation for our project was to “democratize” the AR authoring process so that anyone, regardless of their ability to code, animate, or create 3D models, could find success in producing an AR experience. We defined our problem statement as, “How might we envision a "no-code” augmented reality creation platform for all?”

Competitive Review

The competitive review we conducted included web-based applications that required no additional hardware, wearables, plug-ins, or round trips to third-party applications to utilize. We carried out our analysis based on three factors: 1) Ease of Use; 2) Quality of Use; and 3) Benefit of Use. We documented the relative strengths and weaknesses for each dimension and then synthesized the best features into our platform and made note of those to avoid.

Some of the strengths we included were:

• Access to “drag and drop” pre-made asset library elements
• Start-up templates
• Distilling projects into a discreet number of steps
• integration with other software and hardware

Weaknesses that we sought to avoid included:

• Overwhelming interfaces with too many choices
• No guidance as to what to do next
• Lack of integrated help
• Lack of visual cues
• Inability to go back to a previous screen or “undo” a prior action
• High learning curve for the first-time user
• Requiring a secondary app to generate a QR code
• Lack of precision when positioning elements

We further identified new features such as voice guidance to help navigate users through the authoring process, and streamlined the elements, properties, interactions, and templates to a manageable number.

Guiding Principles

We believe in having an AR platform that is a simple standalone “all-in-one" application. Our guiding principles included that our platform is inclusive and accessible, giving the user the ability to create content regardless of their education, technical expertise, age, first language or access to advanced technology. Additionally, our team focused on simplicity, and ease of use and envisioned our product to provide onboarding guidance, encourage users to track progress, and minimize the users' cognitive load. We decided to prototype our platform as a web-app because it provides universal access.

Methods for this phase

Method 3: Exploratory Interviews

Goal (contribution to the project):

Conduct live interviews to understand the goals and challenges of any user who wants to create and share an AR experience with someone else. Interview findings will be used to develop design implications and shape our product direction.

Detailed Method Description:

Our team recruited 9 participants from the CDM participation pool and through personal connections for our live user interviews. Before the interview, participants were sent an informed consent form and a screener questionnaire. The form asked them to confirm that they were 18 years old or older and use a 5-point Likert scale to indicate their awareness of AR, and their level of familiarity with creating AR content. We identified participants as either an “expert” or “novice” based on these responses. Four out of 9 participants were identified as expert AR users, and 5 out of 9 participants were novices. Each team member interviewed at least 2 participants remotely via Zoom.

Interview Protocol Overview

Warm-up

Our warm-up section contained a separate set of questions for the novice and expert users. Questions for the novices were about their general knowledge of the AR world. We asked our experts to share their thoughts about AR platforms, their process when making AR content, joys, limitations, and suggestions to improve AR creations.

Deep Focus

In the “deep focus” section of the interview, we asked participants to walk us through their process of showing someone how to perform an activity. This could be any form of activity, but also be related to their hobby. The deep focus section of our protocol also involved showing the participants videos based on the level of familiarity with creating AR they indicated on the questionnaire. If they indicated 1 or 2, we then shared an external link to a video that showed a demonstrative assembly of IKEA furniture using an AR app called AssembleAR. After they watched the video, we asked the participants about their impressions as on IKEA AR experience and what they feel about creating such content.

If the participants indicated a 5 or 6, we gave them a scenario to imagine creating a graduation invite for their brother and asked them to explain their process and the tools they would use. We then shared a link to a video from an app named iGreet that showed a birthday greeting card. After they watched the video, we asked the participants to share would they change about the invite, what was challenging and what was their overall impression.

Wrap-up and Demographics

We ended our interviews by asking a few demographics questions relating to participants’ occupation, gender, and ethnicity.

Full Interview Protocol

Coding the transcripts:

We recorded the user interviews and transcribed them using Otter.ai. Then, each team member uploaded the transcription to Atlas.ti and open-coded the interviews according to what we felt was relevant. Coding triggers included demographic information, negative words such as “difficult, constraining, confusing” and positive words such as “like, helpful, useful.” We also captured comments and quotes related to their hobbies, daily activities, tools, suggestions for features to include in our platform, and perceptions of what they felt was missing from current AR tools. We also inquired about their motivations for using AR, and how they would use AR to enhance the experience of their hobby.

We exported our Atlas.ti codes to Excel, and then to a “FigJam” board to analyze and organize the main themes from the interviews. This allowed us to collaboratively work on the platform to create an Affinity Diagram. The discussions during that session helped us come to conclusions about our personas and ideas for our card sort categories and cards.

Interview Participants

We recruited nine participants through convenience sampling and the CDM Participant Pool. Note: *Participants were designated as “Novice” users if they stated that they had never made AR content before.

Participant	AR Knowledge	Age Range	Occupation	Gender	Ethnicity	Hobby
P1	Novice	18-24	Student	Male	African American	Working out
P2	Novice	18-24	Secretary at Tennis Club	Male	Caucasian	Playing games with friends
P3	Novice	18-24	Student	They	Asian	Walking
P4	Expert	45-54	3D Modeler/Engineer	Male	Caucasian	Beer making
P5	Novice	35-44	Learning Specialist	Female	Caucasian/Hispanic	Music
P6	Expert	18-24	Student	Female	Biracial	Figure Skating & Social Media Content Creation
P7	Novice	25-34	Strategy Manager at a bank	Female	Asian	Walking
P8	Expert	25-34	Staff Product Designer; UX Technologist & VR/AR Enthusiast	Male	Caucasian	Photography, social media
P9	Expert	45-54	UX/UI Designer	Female	Caucasian	Walking, Pilates, reading books, watching Netflix

Diversity and Accessibility Considerations

Our team took several steps to include diverse participants based on age,level of experience with augmented reality, English as a second language, learning styles, vision impairments, hearing loss, access to latest technology, and whether they were “digital or non-digital natives”. First, we included definitions of augmented reality and content creation tools in our live interview and prototype evaluation scripts to ensure a common understanding of the domain area. During the live interview phase, we spoke with a participant in her fifties who started her career in photography and came to technology much later. That participant took a 6-week course on AR and expressed challenges with learning the ZapWorks software due to how “buggy” it was and that it required the assistance of an instructor to provide “workarounds” to complete tasks. From that interview, we wanted to ensure that our platform was intuitive and included integrated Help and guidance. We also interviewed an experienced AR enthusiast in his thirties who suggested that allowing the user to “tinker” with existing 3D models and pre-made templates would help the novice user “see what’s possible” and make them feel successful right away. He expressed that his learning style is to “take things apart” to see how they are made and “deconstruct stuff to understand it better.” Based on his feedback, we included several “starter” project templates, pre-made elements, and a gallery of inspiration to make the novice user feel confident and support the learning style of “reverse-engineering.”

Interview Findings

Based on our 9 interviews, we identified 2 primary users: Novice and Expert.

Novice

Five out of 9 participants (P1, P2, P3, P5, and P7) of our live interviews were categorized as “novice” users. All 5 novice users indicated “1- not at all familiar” on the 5-point Likert-scale question about their familiarity making AR content and they also shared some concerns. One of the participants (P2) said, “Getting used to how to use an app with AR capabilities.” Another participant (P5) specified that “some people may only be able to afford the bottom tier, they may not have unlimited data, there may be things that happen beyond our control, like power outages, or, or other things where internet connection suddenly gets severed.

There are certain requirements of peripherals in order to bring the augmented reality to life which again, is and not only a financial barrier, but certain tools cognitively cut with cognitive consideration, certain age groups can't do it.
- Jade

Since we showed novice users the IKEA furniture video, all the participants shared impressions that were both positive and negative. Three out of the 5 novice users (P1, P5 and P7) found the video to be interesting. Two of participants (P2 and P3) found it difficult to connect with it, with one (P2) mentioning that the “instruction manual are easy to follow and having a platform like this might be difficult at first.” Another participant (P3) said, “It’s very interesting that you can hold your camera, I show the tutorials, just inside a mobile device, but then I feel that I just feel it’s more like the same, like I’m watching an animation rather than augmented reality.”

Expert

Four out of 9 participants (P4, P6, P8, and P9) of our live interviews were categorized as “expert” users. All 4 “expert” participants who have used AR platforms in the past suggested some improvements that they would like to see in an AR platform. One of the participants (P4) suggested having tutorials, a “Preview” feature, ability to see the programming side of the platform, and better file management. Another participant (P8) suggested that a AR platform should provide webinars, customer support and examples of what is possible.

Additionally, the expert participants also shared the frustrations and limitations they faced while making AR content on other platforms, which also confirmed some of the drawbacks we found during our competitive analysis. One participant (P4) said that “additional add-on was required, and it was a content specific platform so unable to adapt.” They also mentioned that the software required frequent troubleshooting, has limited instructions, and has an unfriendly user interface.

If you're trying to make something even for your project, like if you're thinking how do we show something that's a social AR experience, I would say a big piece of that will be figuring out if multiple people are able to see the thing and interact with it.
- Ryan

Insights

Considering the thoughts and impressions shared by the participants, we organized our interview findings into 3 main categories 1) Ease of Use, 2) Simple Onboarding, 3) Functionality.

Ease of Use:

All 9 participants highlighted having an easy-to-use AR platform that has easily recognizable UI elements and is simple to navigate to effectively create AR content. Expert users utilize an AR based platform to create content that relates to their hobbies such as painting and photography, while novice users expressed interest in exploring and creating AR content. However, both types of users stated that not having the right features in an AR platform makes it challenging to create effective AR content.

Simple Onboarding:

Seven out of 9 participants said that they would prefer to have an AR platform with simple onboarding and “help” features. One participant said, “With AR you don't want to have like a whole lot of stuff. You want it to be simple.” Four out of 9 participants wanted feedback on where they are in the process when creating AR content. One participant even mentioned that a “blank canvas can be intimidating” and suggested, “helping the user understand what is even possible and help them visualize the steps.”

Functionality:

Six out of 9 participants highlighted that since AR is “the next big thing,” the platform’s performance should be seamless and uninterrupted. Participants who had experience using an AR platform, SparkAR, noted difficulty dragging and dropping objects to the workspace and a lag in the system during the content playback. One participant mentioned that when using SparkAR, an Augmented Reality application, it was “sometimes hard to test it out and move objects especially when dragging and dropping them in the scene.”

Based on these findings, we decided to integrate the following features into our system:

Easy navigation to help users find and utilize the main features of the AR platform and create content.

A step-by-step progress bar and verbal guidance help the user progress on the content they are making.

The ability to connect with the community to ask questions whenever needed.

A set of templates would allow the user to start somewhere and create content of their choice. 

Method 4: User Personas

Goal (contribution to the project):

Generate user personas based on the interview responses. This will give us a better understanding of who the “archetype” users could be and empathize with their pain points, motivations, and goals. This will allow us to seek ways to design an enriching and meaningful AR experience within our platform.

Detailed Method Description

After conducting and analyzing our 9 user interviews, our team created 2 distinct user personas in Figma: Olivia, an Occasional Educator (Novice User) and Aaron, AR Content Creator (Expert User). Both personas include a brief biography, a quote, the motivations, frustrations, goals, personality traits, and brands they are familiar with.

Results

We used qualitative interviews to better understand the pain points of expert users when making AR content and derive inferences about what hesitations novices may have towards making AR content. After the interview, we synthesized our transcripts and notes to highlight major themes. The themes and trends we analyzed from the interview data helped us create two user personas: Olivia and Aaron. We highlighted the goals, motivations, frustrations, and personality that we felt were most relevant and that presented opportunities that our platform could address. These personas served as a guide for designing our lo-fi prototype.

If AR was easy enough to create content for, I would use it to teach others because I believe it's going to be the next "big thing" in YouTube."

Bio:

Active on social media and posts about her origami creations.

Motivations:

Interested in educating people about the craft of origami.

Goals:

Looking for a better way to demonstrate origami technique.

Frustrations:

Finds it difficult to create video animations.

I like to tinker and experiment with AR to learn how things are built. I'd appreciate a platform where the community has a centralized way to share what they've made, build upon it, and learn what's possible.

Bio:

Web developer with a passion for photography and AR.

Motivations:

Interested in sharing how to create complex AR creations.

Goals:

Wants to build a library of multi-player AR games.

Frustrations:

Looking to dive deep into the code and have easy previewing.

Affinity Diagrams

After coding our interviews, we organized the data into salient themes and categories based on Expert and Novice users.

Expert User Categories:

• Social Media (Community, Platforms, Behavior, Benefits)
• Activities and Hobbies, including challenges
• Demographics (Personality Characteristics, Occupation, Gender, Ethnicity, Age)
• AR/VR Hardware and Wearables
• AR/VR Software
• Recommendations for AR experience
• AR Benefits and Limitations
• AR Process
• AR Use Cases

Novice User Categories:

• Typical Day Activities
• Tools and Processes for engaging in Activities
• Social Media and Internet Usage
• AR Examples
• AR Challenges
• Recommendations for AR experience
• AR Content Creation, AR Teaching, and Graduate Invitation Scenario
• Impression of IKEA AR Video
• Demographics (Occupation, Age, Gender, Ethnicity, and Personality Characteristics)

Methods for this phase

Method 5: Open Card Sort

Goal (contribution to the project):

Conduct an open card sort activity within our team that will help us decide the overall flow of our platform. This will help us determine the best process that our target users will follow and provide us with the opportunities to decide on card names for our hybrid card sorting activity.

Detailed Method Description

The open card sort allowed each team member to brainstorm ideas on categories for a card sort. We used data from our interviews and competitive reviews to develop the categories. We then added these cards and categories to another FigJam and met to discuss our rationale for the choices made, and how the proposed structure would help both novice and expert AR users move through the application. We transferred our list to an Excel spreadsheet and had a cumulative total of 112 cards, after removing duplicate instances. We consolidated our list of cards to arrive at a total of 64 items (9 categories and 56 cards).

Results

Each team member individually brainstormed navigation options for our platform. A total of 165 navigation options were generated and as a team we put them on the Excel sheet, performed an open card sort, and then discussed which cards and categories are important to include in our hybrid card sort activity. This activity also allowed us to understand the major workflow of our platform and gave us an idea of what features are important to include.

Method 6: Hybrid Card Sort

Goal (contribution to the project):

Conduct and analyze a hybrid card sort for our AR-based platform. How participants categorize items will determine the navigation, category names, and information architecture and hierarchy of the prototype.

Detailed Method Description

Twelve remote participants from our class completed this activity. Using the insights from our design charrette exercise, we identified the primary categories and the cards that would nest within those categories. Optimal Sort was used to create and analyze digital hybrid card sort. Participants were asked to share their responses to the following pre-test and post-test questions:

• Indicating on a scale of 1 to 5, their level of familiarity with creating AR experiences
• Sharing any category that they felt we overlooked and should include in our platform.

Results

The card sort was attempted by 15 participants, out of which three abandoned and 12 completed it. Seven out of 12 participants were not familiar with creating AR content, 3 out of 12 participants were somewhat familiar with creating AR content, and 2 out of 12 participants were moderately familiar with creating AR content.

Participants were asked to categorize 56 cards under eight categories.

1. Assets
2. Camera
3. Edit
4. File
5. Help
6. Insert
7. Tools/Objects
8. View

Participants also created two additional categories, where they placed 4 of the 56 cards. Because the occurrence of cards in addition to other categories was low, we did not consider these categories as a part of our results:

• Unsure
• I don’t know what that is

The highest percentage of agreement achieved among cards was 100%, and the lowest was 8%.

Card Sort Results with Highest Agreement Percentages

The table below lists the cards with the highest percentage of agreement from highest to lowest and associated categories participants placed cards into. The cards above matched what our team intended. However, we determined that we should exclude “Add” from cards “Add Text, Add Animations, Add Sounds, Add Audio” as we felt it might be confusing and having it under “Insert” would itself be self-explanatory.

Cards	Categories	Percentage of Agreement
Ask the Community	Help	100%
Customer Service	Help	92%
Record	Camera	92%
Save	File	92%
Add Animations	Insert	92%
Tutorials	Help	83%
Start a New Project	File	83%
Send to Back	Edit	83%
Capture	Camera	83%
Add Text	Insert	83%
Add Sound	Insert	83%
Add Audio	Insert	83%
Grid View	View	83%
Add Picture	Insert	75%
Characters library	Assets	75%
Draw	Tools/objects	67%
QR Code generator	Tools/objects	67%
Scan	Camera	67%
Sound Library	Assets	67%
Undo	Edit	67%
Publish	File	67%
Project Type	File	67%
Import/Export	File	67%

Card Sort Results with Lowest Agreement Percentages

The table below lists the cards with the lowest percentage of agreement; they are organized from lowest to highest, along with the associated categories participants placed cards into.

Cards	Categories	Percentage of Agreement
Backgrounds	Tools/Objects	8%
Sound Library	Insert	8%
Color	View	16%
Project Type	Edit	16%
Publish	Assets	16%
Project Settings	Edit	16%
Group	Tools/Objects	25%
Font	Edit	25%
Artboard size	Edit	25%
Layers	View	33%
Scene Selection	Tools/Object	33%
Assemble	Tools/Object	33%
Starter Template	Assets	33%
Interactions	View	33%
Signup/Sign in	Help	33%
Arrows	Tools/objects	41%
Shapes	Tools/objects	41%
Package	Assets	41%
Buttons	Tools/Object	41%
Preview	View	41%
Video	Camera	41%
Arrange/Organize	Tools/Objects	41%
User Setting	Help	41%
Search	Help	41%
Preview	View	41%

Card Sort Results with Highest to Lowest Agreement Percentages

The overall percentage of agreement among cards and categories was low. The highest agreement rate reached among categories was 42%, and the lowest percentage was 19%. The table below lists the categories by the percentage of agreement from highest to lowest.

Categories	Percentage of Agreement	Number of Cards
Help	42%	12
Insert	32%	18
File	31%	24
Tools/Objects	30%	38
Edit	26%	35
Assets	26%	30
Camera	26%	16
View	19%	26

We determined that the primary reason the agreement among categories was low was that the participants placed so many different cards in each category. The main reasons we identified were:

• Some participants were confused between the categories "Tools/Objects" and "Insert" and they needed more context on what each of them meant.
• Not all our participants were quite familiar with the AR platform, so some card labels were unfamiliar.

The card sort results indicated that we need to revisit some cards and category names and make them more straightforward for our Lo-Fi prototypes. Therefore, based on all the data we collected from this card sort, we made the following changes to our cards and categories to start our Lo-Fi prototypes:

1. We eliminated the “Insert” and “Tools/Objects” categories because we felt they both achieved the same tasks of adding objects or text, or other elements.
2. We split the “Assets” category into sub-categories:
• Elements which included icons, background, buttons, 3D objects, Text, Pictures, Shapes, and Arrows
• Video Library that includes pre-loaded videos
• Sound Library that includes pre-loaded audio
• Tools which include a Pen tool, Brush Tool and Draw Tool
3. We also included an additional category, “interactions,” that will list all the animations and gestures.
4. Additionally, we decided to have a separate “Properties” category that will contain the options to modify properties of any element on the workspace.

Similarity Matrix for the Hybrid Card Sort (Group Activity)

Methods for this phase

Method 7: Design Charrette (Sketching)

Goal (contribution to the project):

Conduct a design charrette to collaborate and quickly explore and generate ideas for the layout, navigation, organization, and structure for the major screens.

Detailed Method Description

After analyzing the card sort results, our team began sketching out the basic lo-fi prototypes. The team decided to use Figma for its collaboration capabilities that allowed everyone to work simultaneously on the prototype. Prior to designing the screens, each team member explored similar AR applications using online creative platforms such as Dribbble and Behance. We then individually sketched ideas using paper and pencil and Procreate (Tablet App). We discussed which screens and features were essential and listed what we needed to evaluate in our usability test session. The design charrette allowed each team member to brainstorm ideas by sketching design concepts separately then coming back to share with the team. We discussed the pros and cons of each approach.

Results

We combined the best ideas from the sketches to inform the Lo-Fi prototype design.

Sketches and Initial Wireframes

Each team member individually sketched layout ideas for our platform then explored translating them to wireframes.

Methods for this phase

Method 8: Lo-Fi Prototyping

Goal (contribution to the project):

Create a low-fidelity digital prototype to showcase the concept, functionality, and prioritized design implications for our AR web-based platform.

Detailed Method Description

We translated each sketch into a separate Figma wireframe. We further refined our concept by iterating based on team feedback. We consolidated the best ideas from each team member's design into one prototype and made it clickable in Figma for our usability testing.

1. Style: Our prototypes were created using a grey color scheme with a font that mimicked handwriting but was easy to read. We used universal icons, a basic logo and a simple, clean layout that was not too overwhelming for first-time users.
2. Features: On our lo-fi prototype, we focused on the following key features:
• Project Dashboard
• Template Options
• Project Workspace
• Assets (Adding Elements, and Text)
• Interactions Tab
• Preview Mode
3. Information Architecture: We used the results from our hybrid card sort to organize the menu names and structure for our platform. Overall, our team created 36 prototype screens and made them clickable for our usability test sessions. We generated a shareable Figma link for our participants.

Lo-Fi Wireframe Prototype

We created a lo-fi clickable prototype for usability testing, targeting the Laptop as the main authoring device with a smartphone for sharing and previewing the AR experience.

View Figma Lo-Fi Prototype

Method 9: Mid-Fi Prototyping

Goal (contribution to the project):

Design mid-fi prototypes and incorporate feedback from lo-fi evaluation to have an improved and intuitive platform.

Detailed Method Description

Applying the findings and results from our lo-fi usability test sessions, we designed our mid-fi prototype. Instead of using Figma, as we had done previously for our lo-fi prototype, our team decided to use Adobe XD for its easy voice integration. Adobe XD has a built-in text-to-speech playback feature that allowed us to easily add text which produced system generated speech. At first, we tried using Figma for our voice-guided prototype but had to install a plugin called “Anima” which was confusing, cumbersome, and time-consuming to use because it required embedding code for an mp3 file to play the verbal guidance.

For our mid-fi prototype comparison evaluation, we needed to produce a total of 106 screens (53 screens for Prototype A – Without Voice and 53 screens for Prototype B – With Voice Enabled).

Process:

We created an Adobe XD file, revised our logo, and added color to the design. The screens were then made clickable using the “Prototype” feature of Adobe XD. We quality tested the two prototypes to ensure all features functioned properly before usability testing.

Design Changes:

We updated the layout, visual style and organization of content and navigation on our platform to better meet the needs of the user. Since we had two prototypes, the significant design changes took place on Prototype A which did not have verbal guidance. Prototype B, which included the verbal guidance feature, was a copy of Prototype A, and we only made a design change by adding the “Steps” feature at the top.

• Color Contrast: For our mid-fi prototype we used a monochromatic color palette with an accent color. We ensured that the colors had a high contrast, and the text was easy to read. We tested our color palette using a software called Stark, a contrast-checking plugin for Adobe XD
• Logo: We decided to name our platform “ExplorAR” and designed a logo that was based on the branding and colors used in our overall platform.
• Elements: Since some users had difficulty differentiating between the terms “Assets” and “Tools” on our lo-fi prototype, we changed the “Assets” label to “Elements.” It contains icons, backgrounds, objects, text and drawing tools, photographs, video, and sound clips.
• Split Preview and Publish as Two Different Steps: Participants from our lo-fi usability testing were confused with having “Preview and Publish” as a combined step, and recommended splitting it up into two separate steps. Therefore, in our mid-fi prototype, we split them into two separate buttons. A QR code and link is generated to be shared with others once the participant clicks the “Publish” button.
• Prototype B changes: Our second prototype had all the same changes as our first mid-fi prototype. The only change we made was having a “Play” button in the “Step” feature at the top that indicated clicking on it will play a voice that will guide the user on what step they are on.

Mid-Fi Prototype

We created two mid-fi clickable prototypes for A/B Comparison and usability testing, targeting the Laptop as the main authoring device with a smartphone for sharing and previewing the AR experience. Please note: not all functionality is wired or clickable. To see all the screens, view the carousel below.

View Adobe XD Mid-Fi Prototype Version A (without sound)

View Adobe XD Mid-Fi Prototype Version B With Sound (Note: Volume is loud, adjust as needed).

Methods for this phase

Method 10: Lo-Fi Prototype Evaluation

Goal (contribution to the project):

Conduct a lo-fi formative evaluation to identify potential usability problems our users may encounter when trying to complete tasks. This evaluation feedback will allow us to modify the prototype early in the design process, test ideas, and collect feedback on revised designs more efficiently.

Revisions or adjustments:

We initially planned to have 5 participants and only tested 4 participants.

Explain changes:

The reason we tested 4 participants instead of 5 was because of the extra effort required to design a quality initial prototype.

Detailed Method Description

We recruited 2 participants through the CDM participant pool and 2 from personal connections for our lo-fi prototype remote moderated usability testing. The only in-person session was conducted on our 89-year-old participant who required assistance accessing the prototype. All other sessions were conducted through Zoom and were video recorded after receiving the participant’s verbal consent. Before conducting the test sessions, our team created a usability test protocol that included a brief introduction, a participant's statement of verbal consent, background questions, warm-up questions, definitions of augmented reality and content creation, a scenario, four tasks that participants were asked to complete, and demographic and wrap-up questions.

During the sessions, we encouraged the participants to think out loud while performing the tasks. If the participant asked for help, we provided directions but did not give them a direct path to complete the task. We evaluated how well participants could navigate and perform basic tasks such as starting a project, selecting, and editing elements, previewing, and publishing a greeting card. We specifically chose a greeting card because of its universal qualities as most people have either received or sent greeting cards. We deemed evaluating a DIY task as too complex for this phase of our project.

Each task had a post-task question, where we asked the participants to rate their confidence level of completing the task using a 5-point Likert scale (1 = Not at all confident and 5 = Extremely confident). Each task was measured as Pass/Fail and we also noted the time it took for each participant to complete the task.

Lo-fi Interview Questions and Test Plan

Lo-Fi Prototype Goals

Task	Name	Goal
1	Create an Account	Determine if our login page is easy to understand
2	Start a new project using a greeting card template	Determine if the new project creation process is easy and if our prototype has meaningful buttons, icons, naming conventions, and architecture
3	Make modifications to the template	Determine if the step-by-step approach makes sense and if the layout of our prototype is easy to use
4	Review the Preview Mode	Determine if there is a clear distinction between “preview” and “publish” in our prototype

Lo-Fi Usability Test Participants

We made note of how familiar participants were with augmented reality using a 5-point Likert scale.

Participant	AR Awareness*	AR Creation	AR Familiarity**	Occupation	Age Range	Ethnicity
P1	2	No	2	Housewife & Artist	65-90	Caucasian
P2	1	No	1	Student	25-34	Asian
P3	5	Yes	5	Student	25-34	Asian
P4	1	No	1	Gemologist	34-44	Asian

*On a scale of 1 to 5, how aware participants were of Augmented Reality (AR) experiences (1 is not at all aware, 5 is extremely aware). **On a scale of 1 to 5, when 1 is not at all familiar and 5 is extremely familiar, what was the participant's level of familiarity with creating AR experiences.

After completing the lo-fi usability testing, our team compiled the results and added the participant demographic information, insights, notes, time on task, confidence rating and other important information on a Microsoft Form. The team then discussed the findings and made a list of what changes are needed for our mid-fi prototype.

Diversity and Accessbility Considerations

During one of the lo-fi prototype evaluations, we moderated an in-person interview with an 89-year-old artist with monocular vision due to suffering an eye stroke in one eye. The moderator controlled the keyboard, while the participant pointed where to navigate on the screen. This participant has peripheral vision in one of her eyes and has other age-related natural declines in memory and hearing. She indicated that a verbal guidance audio feature would help her know which steps to follow when completing a task. As an artist, she also expressed a desire for hand-drawing tools such as brushes, pens and pencils. Both features were included in our mid-fidelity prototype. Because she uses an iPad, Chromebook, and flip-phone we felt that having a desktop-sized platform would offer the greatest flexibility for authoring, as older participants may experience fatigue if using a gesture-based application, or one requiring a smartphone.

Results

Four remote usability test sessions were conducted, where the participants performed four tasks.

Task Completion Rate

To evaluate the effectiveness of the prototype features, we rated each completed task as “pass” or “1” and each failed task as “fail” or “0.” We then calculated the percentage of success.

Task	Name	Completed
1	Create an Account	3/4
2	Start a new project using a greeting card template	4/4
3	Make modifications to the template	1/4
4	Review the Preview Mode	4/4

Average Task Completion Time and Confidence Ratings

The table below summarizes the average completion time for each task and the confidence rating.

Task	Name	Average Time on Task	Confidence Rating
1	Create an Account	02:09 minutes	2/4 rated task as 5 - extremely confident
2	Start a new project using a greeting card template	00:47 seconds	3/4 rated task as 5 - extremely confident
3	Make modifications to the template	06:40 minutes	2/4 rated task as 2 - somewhat confident
4	Review the Preview Mode	03:20 minutes	3/4 rated task as 4 - slightly confident

Summary of Participant Feedback for Lo-Fi Prototype

Our participants expressed both positive and challenging experiences while using the platform.

We categorized feedback based on the following 8 dimensions:

1. Wayfinding

Locating Elements

Areas of Opportunity

In the “Make modifications to the template”   task, four out of 4 usability test participants had difficulty locating the maple leaf element. They highlighted that the overall prototype flow was somewhat confusing.

Orientation and Progress Steps

Positive Feedback:

P2 remarked about her positive experience when editing the card, “Everything is pretty. I liked the progress bar and the links to ‘Look Inside, Front, Back, Go to Edit Mode.’” P3 appreciated the numbered steps and said, "The stepper is there to guide me though the steps and where I am in the process.”

Areas of Opportunity:

While editing the greeting card, one participant stated they wanted to see the progress they had made and understand “where they are,” and how many more steps they need to complete the card. P3 suggested, “I want to see the progress I have made if this is a step-by-step process. Such as ‘there are 2 more steps after this.”

2. Freedom and Flexibility

Customizing the Greeting Card

Areas of Opportunity

During the “Make modifications to the template”  task, when asked about what additional information participants expected to see when customizing their greeting card, three out of 4 participants expressed having more choices on which elements to add to the card. P1 also wanted to choose from a variety of greeting card “types” (e.g., birthday, graduation, and other important milestones). P1 said, "You chose a heart (icon). Well, that’s what bothered me. I thought that I had to make a card with a heart. I was expecting to choose from the types of greeting cards.”

When instructed to add the Maple leaf icon to the card, P1 remarked, “I might want 3 maple leaves, but I don’t know how to do that?” She was also unsure if she could add text and images in places other than where the placeholders in the starter template indicated. P1 wanted the freedom to put elements where she wanted. Additionally, P1 stated, “Every step I have choices of what I want to do for each step. I feel I did not have that here.” Also, P3 stated, “There is no way for me to know that I can add more text, so I want to be able to create a dialog box where I can click where I want, and a text box will come up.”

P1 stated she wanted to draw her own elements on the card. She recommended, “Tools to draw would be very good. I would use something like that to make up my own flower and branch. My own things.” P2 expressed confusion about using a pre-made template design. She shared, “The screen is easy and straightforward, but I was confused as to why the card already has elements on it like the heart.”

Based on previously using tools like Canva and Wix, P3 recommended, “Add a ‘Replace’ button to replace heart with the maple leaf. Clicking on the Icons would open up a sub-pane to show the different icons.”

Assistance, Onboarding and Help

Areas of Opportunity:

All four participants asked for some form of assistance during the usability test when completing a variety of tasks. While performing the “Make modifications to the template”   task, some participants expressed a desire for additional help when tasked with adding a maple leaf icon to the card. P4 said, “I felt the task was challenging, as I am not a techie person. I would have liked to have more information and guidance on this topic. The buttons were not working, and I have never seen something like this before. It was overwhelming to know what to do.”

3. Starter Templates & Sample Projects

Positive Feedback

In the “Start a new project using a greeting card template”   task, P1 expressed interest in the Video Library, “I could go to the Video Library and see what they have done with various cards – like samples.”

4. Terminology and Labels

Areas of Opportunity

In the “Review the preview mode”   task, some participants expressed confusion that they did not understand what certain labels meant. P3 expressed confusion about the Preview and Publish button labels. He stated, “The wording ‘preview and publish’ confused me on whether it published because the ‘step’ seems as if it’s a button.” P1 thought that “Publish means Print me Out.” During the “Make modifications to the template” task, P1 did not understand the term, “interactions.” She said, “I’m not sure if interactions mean it’s a moveable card.” P1 also did not understand the term, “Assets.” She remarked, “I don’t know what Assets is. I have no idea what that means.” P1 was confused about abbreviations. She stated, “I don’t know what ‘px’ stands for. I would expect the size.” P2 stated that there could be a better way to phrase the ‘Tools’ label under the ‘Assets’ tab. They also suggested simplifying the “Upload Assets” button to just ‘Upload.’

5. Organization and Hierarchy

Positive Feedback

In response to the “Make modifications to the template”   post-task question, when asked what additional information participants expected to see when asked to add interactions to the maple leaf, P3 stated, “Location for Interactions makes sense because I’m used to it, but for the layman, I would prefer it at the bottom – similar to how it is in Mac.”

Areas of Opportunity

In the “Start a new project using a greeting card template”   task, when asked what participants thought they could do upon arriving at the project workspace, some participants expressed confusion and did not know what to expect under certain sections. P2 stated he was unsure what the Tools section would consist of. In the “Review the preview mode” task, when asked what participants expected to happen after reviewing the preview page, P2 suggested a different placement for the Publish button and recommended moving the publish button somewhere near the progress bar at the end. They shared, “I feel this is a natural way to complete and notice that step.”

6. Authoring with Confidence

User Control

Positive Feedback:

During the “Review the preview mode”   task, when asked what participants thought would happen next after previewing the page, P4 stated, “It was clear that we could print, email, or send the card and review it on multiple devices like phones and tablets.”

Areas of Opportunity:

During the “Make modifications to the template”   task, some participants did not understand what they were taking action on and what to expect. P3 said, “Is the entire scene for which I’m adding interactions to? Or is it to individual elements? Some goes for Properties.” When asked if there was anything participants think we should add to the experience, P3 stated, “Have only the interactions and properties associated with the elements open up, rather than everything displayed at all times. A separate pop-up should come up when user selects something on the scene, so they know that these are the interactions/properties available to add for that scene/object.” He also noted, “The Adding Interactions step did not make sense because all 3 properties have different functions. I think you should segregate them out so that interactions and properties are a pane of itself.”

P3 expected to see a “play” button so that the interactions that he has selected would be played. In the “Review the preview mode” task, when asked what participants thought would happen next after previewing the page, P3 also remarked, “When I clicked on the Preview and Publish, the entire left pane went away, so I’m not sure what happened. I’m in a limbo wondering did it was published or not? I’m assuming it did not publish.” He also stated, “AR won’t work with desktop so users should be informed that they would need to open on mobile and provide a QR code for them to scan.”

7. Preference for Verbal Guidance

Positive Feedback

After completing all 4 tasks, we asked our participants if they preferred having verbal guidance on our platform. Three out of 4 participants indicated that they would like to have this feature. One participant (P4), however, remarked,”This is new to me so I would not know what to add but more guidance would be helpful.”

8. Benefits of Use

Positive Feedback

In our post-test debriefing, P4 said that “I would love to share a card with my parents who are far away with my voice on the card. I think they would love that.” Also, 2 out of 4 participants indicated that they would use our platform to create and send an e-greeting card to their friends or family members because it allowed them to have a more personalized experience. When asked how likely participants were to use this product compared to sending an e-card or a physical greeting, three participants stated that they would use it. P1 said, “It could be kind of neat – it would be ‘my thing’. This would be my greeting card that I had made. Only in that aspect would it be fun to be able to do this. No one else. You can’t buy it in a store.” P4 said, “I have not sent an e-card in a very long time, and I feel I would use it if it was easy to use.”

Positive Feedback:

P2 remarked about her positive experience when editing the card, “Everything is pretty. I liked the progress bar and the links to ‘Look Inside, Front, Back, Go to Edit Mode.’” P3 appreciated the numbered steps and said, "The stepper is there to guide me though the steps and who me where I am in the process.”

Areas of Opportunity:

While editing the greeting card, one participant stated they wanted to see the progress they had made and understand “where they are,” and how many more steps they need to complete the card. P3 suggested, “I want to see the progress I have made if this is a step-by-step process. Such as ‘there are 2 more steps after this.”

Method 11: Mid-Fi Prototype Evaluation

Goal (contribution to the project):

Evaluate the mid-fi prototype’s effectiveness and identify the satisfaction level of participants with the features by utilizing moderated user testing.

Detailed Method Description

Our team recruited four participants from our connections for mid-fi usability testing. All four sessions were conducted through Zoom and were video recorded after receiving the participant's verbal consent. Before performing the test, our team created several artifacts. We created a usability test protocol that included a brief introduction, a participant's statement of verbal consent, background questions, warm-up questions, definitions of augmented reality and content creation, a scenario, five short tasks for the first mid-fi prototype and three questions to measure impressions of the second mid-fi prototype. Also, the protocol had demographic and wrap-up questions at the end.

We encouraged participants to think aloud while completing tasks. If the participant asked for help, we provided directions but did not give them a direct path to follow to complete tasks.

Usability sessions began with background questions where we asked participants to choose their awareness and familiarity levels with AR content on a 5-point Likert scale. If they responded with a “1” or “2” on the scale, we played them a short video from “iGreet,” which is an app for making AR greeting cards.

Mid-Fi Prototype A (Without Verbal Guidance)

We utilized the greeting card scenario from our lo-fi usability testing. The tasks for our mid-fi Prototype A were the same, except we broke them into subtasks to make it manageable for the participants to explore our platform. Each task had a post-task question, where we asked the participants to rate their confidence level in completing the task using a 5-point Likert scale (1 = Not at all confident and 5 = Extremely confident). Each task was measured as Pass/Fail, and we also noted the time it took for each participant to complete the task.

Mid-fi Interview Questions and Test Plan

Mid-Fi Prototype Goals

Task	Name	Goal
1	Create a greeting card project	Determine if a new project creation process is easy.
2	Add a “Maple Leaf” and text on the card	Determine if our prototype has meaningful buttons, icons, naming conventions, and architecture.
3	Add content inside of the card	Determine if it is easy to edit inside the greeting card.
4	Add "slide to right” Interaction to the card	Determine if the interactions tab is appropriate and if the participants can visualize the animations that were added to the card.
5	Preview and publish project	Determine if there is a clear distinction between “Preview” and “Publish” step in our prototype.

Mid-Fi Prototype B (With Verbal Guidance)

Playing the verbal feature of Prototype B required an Adobe XD account. To eliminate placing the burden of account verification on our participants, we the facilitators shared our screen and followed completed the login process. We asked participants to explain what they saw on Prototype B’s blank greeting card template screen.

If participants did not mention the “steps” feature at the top of the screen, we directed their attention to it, and asked what they expected it to do.

Verbal Guidance Feature

We asked the following questions to understand if participants found the verbal guidance feature useful or not.

• What do you think about having verbal guidance in the prototype?
• Do you think it will be useful for you when using the platform?
• Looking at this interface, do you think it is clear enough that you can move back and forth between the steps?

At the end of our usability test session, we asked each participant to share feedback about which prototype they found easier to use. We also asked our participants to take a short survey at the end that consisted of Product Reaction cards. This survey was inspired by the toolkit that was developed by Microsoft to understand the “illusive, intangible aspect of desirability resulting from a user's experience with a product.” (Barnum, 2011).

After completing the mid-fi usability testing, our team compiled the results and added the participant demographic information, insights, notes, time on task, confidence ratings and other important on a Microsoft Form. The team then discussed the findings to identify opportunities for future improvements.

Mid-Fi Prototype Usability Test Participants

We made note of how familiar participants were with augmented reality using a 5-point Likert scale.

Participant	AR Awareness*	AR Creation	AR Familiarity**	Occupation	Age Range	Ethnicity
P1	4	Yes	4	UX Designer	25-34	Asian
P2	2	No	2	UX Designer	35-44	Caucasian
P3	3	No	2	UX Designer	25-34	Asian
P4	3	No	1	Animator	35-44	Asian

*On a scale of 1 to 5, how aware they were of Augmented Reality (AR) experiences (1 is not at all aware, 5 is extremely aware) **On a scale of 1 to 5, when 1 is not at all familiar and 5 is extremely familiar, what was the participant's level of familiarity with creating AR experiences.

Diversity and Accessibility Considerations

Our mid-fidelity prototyping evaluations included 4 participants who spoke English as a second language. Two appreciated the verbal guidance because they found it helpful in knowing where they were in the creation process and commented that it would help the infrequent user get re-acclimated to the platform. One suggested that an onboarding feature upon launching would help the new user get familiar with the platform’s features and functionality. Based on our users’ need for clear, concise language, we targeted a sixth grade reading level for our content and used simple terminology for labels and typographic elements.

Results for Mid-Fi Prototype

Task Completion Rate

We asked the participants to complete five short tasks for our first mid-fi prototype. To evaluate the effectiveness of our prototype features, we rated each completed task as “pass” or “1” and each failed task as “fail” or “0.” We then calculated the percentage of success.

Task	Name	Completed
1	Create a Greeting Card Project	4/4
2	Add a “Maple Leaf” and Text on the Card	4/4
3	Add Content inside of the Card	4/4
4	Add "Slide to Right” Interaction to the Card	4/4
5	Preview and Publish Project	3/4

Average Time on Task, Ease of Task, and Satisfaction Ratings

We also noted the time the participants took to complete each task and their rating on a scale of 1 to 5 on how easy or difficult the task was and their satisfaction level for each. The table below summarizes the average completion time for each task, ease of task rating and satisfaction ratings.

Task	Name	Average Time on Task (m:s)	Ease of Task Rating	Satisfaction of Task Rating
1	Create a Greeting Card Project	02:25 minutes	4/4 rated task as 5 - Very Easy	4/4 rated task as 5 - Satisfied
2	Add a “Maple Leaf” and Text on the Card	00:52 seconds	3/4 rated task as 4 - Somewhat Easy	2/4 rated task as 5 - Satisfied
3	Add Content Inside of the Card	00:21 seconds	4/4 rated task as 5 - Very Easy	3/4 rated task as 5 - Satisfied
4	Add "Slide to Right” Interaction to the Card	00:40 seconds	2/4 rated task as 4 - Somewhat Easy	2/4 rated task as 5 - Satisfied
5	Preview and Publish Project	01:23 minutes	4/4 rated task as 5 - Very Easy	2/4 rated task as 5 - Satisfied

Summary of Participant Feedback for the Mid-Fi Prototypes

Our participants expressed both positive and challenging experiences while using the platform. The following is a comparison of both Prototype A and B, in which user preference was equally split.

We categorized feedback comparing Prototype A and Prototype B based on the following dimensions:

1. Wayfinding

Prototype A

Orientation and Progress Steps:

Two out of 4 participants had some challenges while completing the “Add a maple leaf and text on the card”   task. One of the participants (P1) said, “I initially thought it was a progress bar. I am adding elements and then adding interactions. I thought I could go back to steps if I wanted, and I could see what I did in a step.” Another participant (P2) found the button in the right corner confusing, and she thought it was clickable and suggested making it more visible. She said, “Saw the button at the bottom to go to the next step and was confused about why it was there.”

Prototype B (verbal guidance)

Orientation and Progress Steps:

We asked our participants if it was clear enough that they could move back and forth between the steps. Three out of 4 participants mentioned that it was not immediately apparent because it specifically lacked visual prominence. One participant (P1) said, “The buttons were not apparent upfront because the buttons were not that intuitive.” P4 said, “I think that's helpful. I just don't think we still need the step number.”

Two out of 4 participants did not click on the “Step 4 Publish” button.

Two out of 4 participants found the placement of some of our icons and tools confusing. P1 and P4 wanted the two-dimensional icons separated out from the 3D objects. P1 suggested putting backgrounds with photos.

Another participant (P4) said, “The steps 1,2,3,4 should be automatic. I shouldn't have to click on them. Once I'm done with a certain step, it should update immediately. At least all these features should be immediately available.”

None of our participants clicked on “Step 4: Publish and Preview” button for the “publish and preview” task. They all clicked on the “Preview” button at the top right corner of screen.

2. Freedom and Flexibility

Prototype A

Assistance, Onboarding and Help

In the “Add a maple leaf and text on the card”  task, one out of the 4 participants (P1) commented, “First I would like to have an onboarding screen where I can understand I can literally get the step-by-step process. I do not know how to add images or text. If there was an onboarding screen, I would have had a more satisfying experience.” P3 highlighted, “Had to get used to the interface and also thought maple leaf is draggable.”

In response to the “Add a maple leaf and text on the card”  post-task question, the concept of “coach marking” was suggested. As defined by the Nielson Norman Group, “Coach marks are temporary messages that educate users through new or unfamiliar product experiences. They can be chained into a sequence to form a tour.” One participant (P1) recommended, “If you click Step 1, it’s telling you to add or edit elements. But I don’t know where those elements are. So, if you highlight those elements at the same time as the recording, the elements will glow to grab the attention of the user. If you go on to Step 2, it will tell you to Add Interaction. But right now, we are on the Elements section, and so I’m not able to see Interactions.” P1 also stated this approach can be repeated for each step.

Prototype B (verbal guidance)

Customizing the Greeting Card

When asked if there was anything we should add to our platform, two out of 4 participants agreed that our platform needs a Z axis in the workspace. One participant (P1) said, “The only thing I would like to add for augmented reality is a virtual environment that is 3D, and for right now I can see this platform as 2D. And so, if you add XYZ axis, the user will understand. Many times, you rotate your objects in a 3D environment and if your greeting card is opened another way it might confuse the user. You need this feature so whenever you rotate the card it will show the axis – it's a property.”

3. Authoring with Confidence

Prototype A and Prototype B

Visual Cues

In the “’Add slide to right’ interaction to the card,”   task, one participant (P1) suggested “show the motion accurately for the motion icons (right now motion is the same for each icon). Have the motion go in the right direction.”

4. Preference for Verbal Guidance

After collecting information about our two prototypes, we asked our participants which prototype was easier to use in our post-test debriefing. We had a split decision among our 4 participants. Two out of 4 participants said Prototype A was easier. Two out of 4 participants said Prototype B was easier.

Prototype A Preferred

One participant (P3) liked Prototype A’s feature but acknowledged some positive aspects of Prototype B. She said, “It would be cool to see that the voice is playing and maybe highlight stuff on what the voice is saying.” Another participant (P2) preferred Prototype A because the second one startled her, and it was not clear to her that she can play and hear a voice. She thought it could be a conflict if someone is visually impaired then they will be using a screen reader. It may be useful, but it would be much more beneficial if we add what you do in the step. She said, “The buttons were not apparent upfront because the buttons were not that intuitive. The steps at the top with triangles seemed like arrows to progress to the next screen and didn't expect that it was a play button, and a voice will play.”

Prototype B (verbal guidance) Preferred

One participant (P1) preferred verbal guidance prototype and said, “This play button was easier, because as I am progressing toward step 3, I can hear the sound. Because I am completing step 1, 2, 3 I know that it is better to have audio scripting behind each one. If the button has my consent – if I am pressing on that button and it is playing sounds, so I don’t think it’s annoying. It’s quite a good feature.” P1 also added, “I think it’s necessary for me because I’m not going to make greeting cards every week and I may have forgotten all the steps, and so I just have someone to tell me and so it’s quite a good feature.” Another participant (P4) slightly preferred verbal guidance because some users may want it. She said, “I wouldn’t say one is easier than the other. I guess it can be useful to have audio, but not a lot of people are going to like having audio if they are working in the library and they don’t have access to a headset. Maybe have a mute option.”

Product Reaction Survey

As a part of our post-test debriefing, we also asked our participants to take a Product Reaction Survey, which included 35 different adjectives and allowed the participants to select any 5 to describe our platform the best. Our analysis highlighted that 3 out of 4 participants (P1, P2, and P3) found that our platform was easy to use, clean and innovative. Two out of 4 chosen participants (P3 and P4) found that our platform was friendly, consistent, and creative. Using this product reaction survey allowed us to learn our participants’ initial impressions of our platform..

Microsoft Product Cards

Participants were asked to pick five cards that best describe their experience with the product.

Comfortable	Creative	Stressful	Ineffective	Friendly
Useful	Difficult	Advanced	Consistent	Inconsistent
Predictable	Exciting	Unpredictable	Effective	Exceptional
Efficient	Innovative	Inviting	Essential	Desirable
Unconventional	Annoying	Helpful	Busy	Confusing
Personal	Clean	Organized	Engaging	Overwhelming
Rigid	Flexible	Easy-to-use	Familiar	Inspiring

Discussion

Overall, we feel that we successfully achieved the goals of our project. We fulfilled our goal of providing an easy-to-use platform called ExplorAR, to help novice and expert users create and share AR-based experiences. Performing a card sort activity helped us in determining the information architecture for ExplorAR. Further, we believe verbal guidance can be helpful for our platform, which confirms our findings from the literature review on AR. Our prototype’s usability testing sessions revealed both the negative and positive effects of having verbal guidance in an AR platform. Our results showed that the tools provided are likely to be prioritized over verbal guidance for an expert user. Our mid-fi prototypes usability test results indicated that our step-by-step feature requires improvements because most participants did not comment on it until we pointed it out to them.

Limitations of Research

There were several limitations to our work on ExplorAR. Our interviews were exclusively conducted remotely on Zoom, and we sometimes faced lag time during the session which caused us to ask the participants to repeat themselves. The relatively short ten-week timeline of this project was another limitation, which forced us to prioritize our efforts. Finally, all mid-fi usability testing participants were personal connections, and we had a couple of UX designers who did the testing, so this limited us in knowing what users from other occupations think about our platform. While we would have preferred to utilize participants that we did not know, given our participant criteria and time limitations, using acquaintances was our only viable option. Our connection to participants could have positively biased some of their responses, as we noted above in relation to our goals.

Future Work

For future work, we would like to interview more participants from different fields to further understand the user's needs and goals. We would also like to interview YouTube or TikTok content creators, to know what is lacking in their field and what can be improved. Additionally, we would like to implement the feedback from our mid-fi usability testing into a high-fi prototype. Because many of our participants interacted with items on the platform that were not actively clickable, we believe our design would benefit from a hi-fi prototype to allow for in-depth exploration. Some of our participants had never used an AR platform, so performing a task on top of learning about an AR platform might have confused our participants. Our future work would have a different set of usability tasks for the novice and a different set for the expert users and this will allow us to be more focused and determine the goals and needs of each type of users.

Conclusion

Our project aimed to develop a "no-code" platform for all types of users (expert and novice) to get inspired by an easy-to-use process for creating and sharing AR content. We leveraged several different methods such as live user interviews, hybrid card sort, and lo-fi and mid-fi prototype evaluations, to help us understand the users' needs, goals, and pain points. Our research revealed that our platform should focus on four main categories: 1) Clear and simple system with easily recognizable elements; 2) Intuitive onboarding; 3) Robust and uninterrupted performance; and 4) Engaging starter templates to foster creativity and support different learning styles.

We incorporated features such as easy navigation to enable users to find and utilize the primary components to create content. Our platform employed design conventions such as a step-by-step progress bar, verbal guidance, and a set of templates to empower users to create content of their choice. We focused our usability testing goals on understanding how our users complete critical tasks related to the designed features. Our lo-fi usability study indicates that we received positive feedback on our prototype. The task failures were minimal and were easily resolved by simplifying the language and the placements of correct icons.

Our second round of usability testing produced excellent results, with an overall 95% success rate for our tasks. Based on our work so far, we believe our platform would be successful in the market as it was well-received by most participants. Further iterations and testing of our platform would help us implement a more refined user experience for our targeted users.