Supporting people living with and around bipolar disorder

A University of Washington HCI+D PROJECT


Helping patients with bipolar disorder find success living independently and support the people they live with.


A visualization system for bipolar disorder patients to manage their symptoms, while helping their roommates validate concerns and connect with the patients larger support network incase of emergency.

Project brief

A self-defined student project over 10 weeks, culminating in a high-fidelity prototype, presentation and UI specification.


User Experience Design, 3D Physical Product Design, Prototyping, Illustration, Information Architecture and User Research.


10 weeks (Sept - Dec 2019)


Figma, Maya, Photoshop, Illustrator, After Effects


3 HCI + Design Students


How Might We...

Support roommates of patients with bipolar disorder, while helping them be better allies to the patient?


Living Around Bipolar Disorder

Bipolar disorder is a lifelong mood disorder, which even effects the people living around the patient. The disordered moods can manifest as depressive, manic or mixed episodes. Many young adults get diagnosed in their early 20's and, while some extreme cases might find it hard to live independently, many patients are able to manage their symptoms through medication and therapies.


Chosen Social Support

Well, now that we are living in an economy where more young adults are finding it necessary to live with roommates, it also necessary for people living with serious mental health conditions to be able to manage their health while living alongside other non-family members.


Supporting Roommates

To get a better sense of all the people impacted by bipolar disorder, My team and I drew up a stakeholder map, This led us to believe that roommates are the closest in proximity to the bipolar disorder patient but least connected to their overall support network. Having decided that this was an opportunity space, we conducted research, ideated and tested our solution to come up with a two-part final prototype.

A custom wearable for the patient


Key Features

Tracking emotional wellbeing.
Patients with bipolar disorder can log their daily emotions via a wearable bracelet that also tracks their sleep and visualizes the composite data on the e-ink surface of the wearable.

Because patients with bipolar disorder want to track their key indicators of mood, emotional state and sleep, as part of their cognitive therapy.

Visual recommendation system.
The wearable is paired with an app that visualizes that patients top level data to the roommate, to validate any concerns they might have about the patients emotional stability, and provides them with actionable recommendations for next steps.

Because roommates of patients want to have more precise information in case of emergency.

the main user interface
interface showing support network

Connect to the support network.
The app will categorize the patients inputted support contacts, using the patient's prioritization, and recommend the best contact to the roommate based on the state of the emergency situation.

Because Roommates of patients want to be connected to the patient’s support system.


Research Overview

Our team conducted secondary research, used cultural probes and conducted follow up semi-structured interviews.

To address our design question, we wanted to understand the dynamic of the bipolar patient - roommate relationship.

We turned to online forums for anecdotal evidence and found very specific descriptions of concerns and frustrations that roommates had living with instability and without context.

“I had no idea that my roommate suffered from psychotic episodes until social workers showed up at my door to take him to the hospital on a Monday.”

- Roommate of bipolar patient via Reddit
“Having someone at home who experiences my emotional and physical state as often as I can, helps me stay on top of when I might be slipping back into depression before a full episode sets in.”

– Bipolar patient via Reddit

primary research

Hearing their Stories

Initially we attempted to reach out to bipolar support groups to find participants but they did not want researchers to interact with their patients so we pivoted to digital probes that we sent out through snowball sampling. We received 5 returned probes.

Through the probes, we found instances of the honest gratitude patients had for their roommates. The semi-structured interviews revealed a much more nuanced live-in relationship. We conducted 2 semi-structured follow up interviews.

Participants: Former roommates of patients diagnosed with bipolar disorder.

Goal: To gain a better understanding of the difficulties associated with living around bipolar disorder.


Key Insights

From these interviews and synthetic artifacts, we used affinity diagramming to synthesize our results into many themes and insights. We focused our three insights on the key elements that described the roommates pain points in the experience of living with the bipolar patient.


Being roommates with a bipolar disorder patient comes with caregiving responsibilities, and most roommates  are willing to take on a certain level of these responsibilities.

“I thought ‘this is hard, I don't know what to do’ not 'this is hard, I don't want to do it'.”

- Participant 1


Roommates want to be connected with patients' support network to better support the patient and to feel supported themselves.

“Being able to communicate with her support network, to be able to communicate with the other people in her life. I wish we could have been more on the same team with her support network.”

- Participant 2


Roommates get most of their information about bipolar disorder from the patient and they want more validated information.

“It was mostly through talking to her and what her experiences were like.”

- Participant 2

design principles

Value Sensitive Design

Using a value sensitive design methodology, based on what we had learned from our subject matter experts and formative research, we developed three design principles to guide our ideation process.

informed consent icon

Informed Consent

All stakeholders in the experience should have a complete understanding of their role and consent to being involved.

a lock icon for privacy


Bipolar disorder is a serious condition and all data involved should be treated securely. Any solution should deeply consider the privacy of the patient and the support network involved.

A sun icon depicting autonomy


In the end, we wanted to give the roommate a sense of control over their circumstances while minimizing burden on them and the patient to know exactly what to do in complex situations.

primary research

Concept Selection

With our research and principles in mind, we generated around 90 ideas and evaluated these concepts. We down-selected by removing all ideas that were related to surveillance or invasion of patient privacy. We created a model based on our stakeholder needs and used that to categorize our remaining ideas into 4 themes - Ethical data tracking, private visualization loop, building a knowledge base and connecting to support network - using dot voting. We ended up with
3 main aspects of a solution experience:
* Track and visualize patients key mood indicators.
* Generate precise information and recommendations for the roommate.
* Help the roommate and the patients network coordinate support.


Why Design a Wearable?

description of wearable


Privacy & Testing

We created wire frames for a potential app to house the data visualizations and tested these ideas through iterative prototype testing,

Since privacy was one of our key principles,
I wanted to create a custom wearable tracker for the patient's key mood indicators.  I wanted to give the patient a discreet experience of being able to log their mood
and sleep at any time of day while tracking their data through a low energy consuming, e-ink display.

Most importantly, closing the data capture and storage loop between the wearable and the app that would only connect via bluetooth.

We tested the wearable interactions with 3 participants.


Insights & Iteration

From the user testing, we were able to make important additions and changes to the prototype that aligned with out design principles. We had spent a lot of time worrying about the consent process but, through user testing, we were able to find an elegant solution that kept privacy and informed consent at the forefront of the design.


Participants wanted a clear description of what their responsibilities would be before they consented to using the app.

Refine: Make a clear and easy consent experience that mirrored interactive messaging.

“I have no idea what responsibility I would have when I agree to join the support network.”

- Participant 1
We made the consent experience look
and feel more like a chat interaction


Participants wanted to review each voice memo before adding an entry, as well as hold down the record button to record. Overall, they hoped the entire process of documenting daily emotions would be quick.

Refine: Give the patients a way to easily input and review their voice memos.

“I don't mind doing this like multiple times every day but I don't want to spend too much time.”

- Participant 2
The wearable has a simple interface and
the memos are saved on the patients app.


In a severe situation, participants are cautious about calling 911. They wish to have alternative recommendations and suggestion before having to resort to emergency help.

Refine: Have the app always recommend an appropriate contact from the patient's support network.

“...text your mom and contact the doctor also, because I don't know if 911 is the escalation that I need right now, based on how bad the situation is and how bad I am able to judge it.”

- Participant 3
Roommates are the closes in proximity to the patient
but least connected to their support network

primary research

Product Journey

From our testing feedback, we realized that there was a gap in understanding how the product would be adopted by the patient, the primary contact or roommate and the support network. Since autonomy was one of our principles, we wanted to make sure that the patient had power throughout the product journey, I created a visual flow to describe how each stakeholder would be added and involved in the experience by
the patient.

Design scope

Understanding Stability

While deciding what metrics to capture, we focused on the warning signs that the roommate actually does see, but is unable to make sense of, when the patient enters an episode. To stay faithful to our principles, I created a figure to understand the minimally invasive data we could capture from the patient to validate the roommates real-world concerns.

My figure outlines the different signs of disordered episodes and which metrics were the least invasive.


Collecting and Visualizing Data

Now that we knew the kind of data we wanted to collect, each of our team worked on different aspects of the product architecture. I focused on the way that the wearable would collect and pass along information from the patient. I wanted to design the wearable so it would only collect minimally invasive data points that would point to emotional stability while connecting to the app locally.  


A Boat on the Water

Moving into the visual design, we put together mood boards based on the research done by our advisor, Dr. Jamie Snyder, regarding how bipolar patients represent their illness. We found several instances of encoding the bipolar experience as a constantly shifting weather pattern on water.

I found inspiration from an animated short called "Little Boat" about a boat in different weather patterns and realized that was more descriptive than a chart or graph as a high level visualization of mood. I sketched out these concepts and our visual designer adapted them into our final prototype.

I created three concepts of boats in different weather patterns to encode the broad bipolar mood situations


A Mood like the Weather

The high level visualizations were based on my illustrations of boats in different weather patterns. The idea is that the roommates would be able to grasp the situation via one quick glance of the app and quickly know what action to take in a potential emergency situation.

I storyboarded three different lived scenarios that roommates might witness
and matched them to the app visualizations we created
Patient and roommate setup the system together.


How It Works

The patient and roommate set up the bracelet and app together.

The patient inputs their preferred support system contact list based on the designations they prefer and consents to share the information with their chosen contact, in this case the roommate.

Patients track their emotions and sleep as part of their therapy.

The patient uses the wearable to log their emotions via microphone and passively track their sleep. This data is visualized on the bracelet as a boat floating in different weather patterns, depending on the overall mood of the patient. This mirrors a cognitive therapy method of tracking emotional fluctuations used to manage bipolar-related mood swings.

Patient and roommate setup the system together.
Patient and roommate setup the system together.

Roommates see the patients overall mood visualized as a boat floating in different weather patterns.

The data gathered by the bracelet is visualized for the patient and the chosen contact, in this case the roommate, via the app. This top level visualization gives the roommate an overall sense of the patient’s mood and short, actionable recommendations of next steps.

The app recommends next steps and the appropriate contact from the patient's support network for hand-off.

The app gives the roommate access to a phone book of the patient's support network, as well as recommendations of which contact to hand the situation off to. This could be a friend, family member or a doctor in case of emergency.

Patient and roommate setup the system together.


Documentation & Video

We created an in-depth video and user interface specification to detail out the experience.




As a team of designers, we realized how important it was to have a strong ethical system.
Throughout this project, we were challenged by situations that made it clear that we could find easier solutions to our design issues if we allowed some of our privacy concerns to slide. At every major design decision point, we were faced with the same kind of challenge and I am so proud of our team for putting our ethical guidelines and value sensitive design principles first. It was important to us to design as though this product would be implemented in the real world by vulnerable patients.


More people have experienced living with bipolar disorder patients than we initially thought.
When we started this project, we thought that we would be dealing with a fairly small population.
As we conducted our research and design, we realized that more and more people came up in our own circles who had these kinds of experiences. It helped us understand that there are many hidden needs that people don't usually expose unless prompted.