PROJECT TYPE
UX Design
(School project)
ROLE
Lead Designer,
Co-Researcher
TEAM
Kayleigh Malik,
David Rezai
TIMELINE
Aug 2023
(3 weeks)
Project Overview
Summary
For a class project, I was tasked with designing a 'fall detection' feature for WatchOS. The target users were those in fall prone demographics, including the elderly. We weren't given the time during this project to do user interviews, but since I was not in the target demographic I wanted to at least research fall statistics and make sure I tested my prototypes on those in the target demographic, especially seniors, in order to minimize any potential biases I might be introducing into my designs.
Problems
Users who have just experienced a fall are under tremendous stress and want an interface that's extremely simple, offers error prevention, provides accessible help options, and is calming. Apple's interface does not allow cancellation options after an action has been taken, only allows users to contact emergency services (which not all users can afford to do), and has an interface that can cause further panic for stressed users through things like the word 'SOS' and prominent red colors.
Solution
I made multiple user flows for serious falls, mild falls, and non-responsive users. Because of high medical costs in the U.S. I also added the option for users to call their emergency contact instead of emergency services, so they could ask for help without cost being a factor. I also added multiple back and cancel buttons and confirmation and loading screens to avoid confusion. Additionally, to help users who need larger fonts, I designed with the largest font size possible from the beginning.
The Design Process
01
Research
02
Ideate
03
Design and Iterate
04
Present
05
Reflect
01 | Research
Fall Statistics
Because this was a project for an introductory UX class, we had not yet covered proper research methodology. Instead, we were given data from the National Council on Aging (NCOA) to inform our understanding of the severity of what fall prone demographics were facing.
30%
Increase in fall deaths from 2007-2016
800K
Annual fall related hospitalizations
3M
Annual fall related injuries in the ER
$101B
Projected medical costs related to falls by 2030
02 | Ideate
Crazy 8s Sketches
As this was a summer class, we didn't have as much time to accomplish steps together in class. The professor had to change up the order of some steps. Originally, the plan was to first create an empathy map and a how might we statement as part of the research step, but due to the semester timeline the ideation steps were moved to the beginning of the project.
Using Crazy 8s sketching, we were asked to think of non-Apple Watch devices that could provide fall assistance, to see if some features could be generated that may have initially been missed when limiting design solutions to solely a smart watch.
Eight hand drawn sketches of non-Apple Watch devices that could provide fall assistance.
I then did a second round of Crazy 8s with more Apple Watch and iPhone-specific layouts and ideas in mind. I initially had the idea to include an additional set of iPhone Health app screens, but ultimately didn't have the time for this assignment and went forward with just Apple Watch designs.
Eight hand drawn sketches of Apple Watch screens that could provide fall assistance.
Context Scenario
Using some of the ideas generated from the crazy 8s scenarios, the next step was to quickly sketch a context scenario depicting how the product might be used and provide benefit to the user.
Hand sketched context scenario depicting a daughter gifting her father a watch will fall detection because she's concerned for his safety.
Empathy Mapping
I created an empathy map with my fellow classmates to determine the pain points and desires of fall prone elderly users. Since we created this empathy map without having interviewed users, my group realized that we were projecting some of our own biases into this empathy map, and we determined that we wanted to test our prototypes on the target demographics in order to minimize this bias.
Empathy map of fall prone users showing what a potential user would think and feel, hear, see, say, and do. There is a summary at the bottom of user pains and gains.
How Might We
With some fellow classmates, I created some basic affinity clusters from the empathy map to brainstorm a how might we statement. This really helped to organize our thoughts and served as an anchor point for the rest of the project.
Brainstorming sheet for creating a how might we statement.
How might we design wearable technology to help the elderly gain confidence, independence, and better access to emergency services?
03 | Design and Iterate
User Flows
A classmate and I brainstormed two separate potential user flows: one for users pre-fall to improve their health and stability and one for post-fall when they need to call for help. My classmate decided to work on designs for the pre-fall user flow while I began working on designs for the post-fall user flow.
After this step, I began working independently on my prototypes and testing.
Two user flows for fall detection: one pre-fall prevention flow and one post-fall help flow.
Low Fidelity Prototype
Using cardboard, paper, and pen, I created a low fidelity physical prototype to begin testing how users responded to the designs. I created two sets of screens to test: normal to mid severity falls and high severity falls.
A series of images showing hand drawn screens on a paper and cardboard watch prototype.
Initial User Testing
I began user testing the physical prototype on friends and family, including my grandparents. Initially, I framed my testing requests as needing opinions on those from fall prone demographics. I quickly learned that this was the wrong approach and made people feel very defensive and embarrassed. By reframing and asking for help making sure my designs were user friendly, people became more receptive.
For testing, I asked participants to show me which buttons they would press, and then I would help move the paper strip to the next screen that they would see.
Pros:
-
Participants felt the design was easy to use
-
Cancelation buttons appreciated
-
The EMS noise alert explanation screen was called helpful
Cons:
-
Participants did not realize the design could use Siri voice commands
-
Requests to add more cancelation buttons, such as when EMS was on the way
-
A few participants expressed concerns over what would (or would not) happen if the watch registered a mild fall but the user was unresponsive
A/B Testing
I then created some initial high fidelity screens, but after realizing how difficult the small screen size of a watch was, I decided to A/B test a few versions of the screen that I had difficulty laying out well.
The main three elements that I A/B tested were the wording (using the term '911' or 'EMS'/'Emergency Services'), button placement (stacked or split screen), and the use of iconography.
Interestingly, elderly users preferred the term '911' due to having grown up with the initial marketing phrasing of, "call 911 if you're ever in trouble" while younger users preferred the terms 'EMS' or 'Emergency Services'. I decided to go with the phrasing 'Call 911' as the elderly would likely be the largest demographic base for this feature.
Users also preferred having iconography and preferred stacked buttons instead of split screen buttons.
A series of images showing the different A/B tested Apple Watch screens.
High Fidelity Prototype
My final deliverable for this project included two user flows.
Mild to Moderate Severity Falls
-
The user is prompted that the watch detected a fall
-
The user is asked if they need help
-
The user can choose to call 911, call their emergency contact (this was added to ease the cost concerns of calling an ambulance), or cancel
-
If users cancel, they're taken to a confirmation screen asking if they still need help (to follow error prevention heuristics)
-
-
The watch then shows a screen showing EMS progress to their location
-
The watch shows a brief confirmation screen when EMS arrives
-
The watch shows a screen stating it will make noise to help EMS find the user
A series of images showing high fidelity screens for low to moderate severity falls.
High Severity Falls or Unresponsive Users
-
If a severe fall is detected, the watch automatically calls 911, but provides the user the option to cancel
-
Alternatively, if a mild fall was detected but the user is unresponsive, the watch will automatically call 911 after a timer runs out
-
If users cancel, they're taken to a confirmation screen asking if they still need help (to follow error prevention heuristics)
-
-
The watch then shows a screen showing EMS progress to their location
-
The watch shows a brief confirmation screen when EMS arrives
-
The watch shows a screen stating it will make noise to help EMS find the user
A series of images showing high fidelity screens for high severity falls or unresponsive users.
04 | Present
Final Presentation
The final day of class was dedicated to everyone presenting their finished projects. The goal of this presentation was to pretend we were presenting to the board of Apple, with the intent of getting our designs adopted and added to the next round of beta testing for WatchOS.
If you click on the presentation cover image below, it will take you the PDF slidedeck of my final presentation.
PDF slidedeck of my final presentation for this project.
05 | Reflect
What I Learned
Asset Management
This was my first large UX project and we hadn’t yet discussed the importance of asset management for case studies. It was difficult to organize everything AFTER having completed the project. I now organize assets as I go!
Time Saving Figma Practices
This was the largest Figma project I had done up to this point, and I didn’t have much time (summer class). I learned about Figma Community UI kits and taught myself components and auto layout to be as efficient as I could.
A/B Testing Utility
Since we didn’t do user interviews for this project, I didn’t realize a lot of my assumptions until testing my A/B versions. I learned so much! This really solidified the importance of learning from the users first for me.
Importance of Minimal Design
Apple Watch screens are TINY, and I deliberately designed in a large font. It was difficult to remove enough for the design to fit the screen while maintaining functionality, but it taught me to look critically at what’s needed.
Next Steps
Were I to continue with this design, there are several next steps I would like to take in order to improve.
Responsive Design
Create different layouts for other Apple Watch screen sizes (38, 40, 41, 42, 44, 45mm) and accommodate layouts for smaller font sizes as well.
iPhone Health App
Create a series of corresponding fall detection screens on the iPhone Health app. (This is already a feature that Apple has, but they also have Fall Detection and I did a project about it anyway!)
Watch Prompt
Design a prompt on the Apple Watch that opens the Health app directly on a users' phone. During testing, many older participants expressed profound confusion (or complete lack of knowledge) of the connection between the Apple Watch and the Health app. I would like to test if a prompt of this nature would help to address that, or if it would make things worse.
Beta Testing (A/B)
I would be curious to see my designs A/B tested in beta with current Fall Detection screens. While I don't believe that my designs are better than those of the Apple team, I would be curious at user reactions to having fewer bright red SOS elements and more choices about who to contact.
