UX Researcher


A mobile app that helps citizens with medical needs prepare and respond to natural disasters.



A mobile app that helps citizens with medical needs prepare and respond to natural disasters.

Illustration by Ketki Jadhav

Illustration by Ketki Jadhav


Project Duration: 7 months
Teammates: Ketki Jadhav, Brooke Sachs, Marina Leao Lucena, Jon Loeb
Responsibilities: UX Research Lead, Interaction Designer


our Client


For my Masters capstone project, my group and I worked for TeleTracking, an industry leading company that has provided hospitals with patient flow solutions for over 25 years. As they enter their next phase, they seek to break into community health and begin offering more citizen-focused solutions. 

Their prompt

Develop a citizen-facing solution that will help alleviate the burden placed on their hospital clients during disasters, who receive a surge of uninjured people seeking shelter in times of crisis. 

Our Focus

Informed by our exploratory research, we decided to focus on citizens with special medical needs (often the elderly) in hurricane prone zones. 

Why hurricanes and citizens with special medical needs?
Hurricanes are the most destructive type of disaster in the United States, both in terms of damage to property and human life. Through our research, we found that those who have a special medical need (for example, dependence on a device such as an oxygen concentrator) are often disproportionately affected by disasters. 

hurricane stats.jpg

In case evacuation from their home is needed, these vulnerable populations are often the ones who will show up to shelter at hospitals, since they require special resources such as medical staff, electricity for their devices, or medication (all of which hospitals provide). 

Key Research Insights

If you’ve been through this so many times, you don’t need to look at lists to tell you what you need.
— Caregiver from Tampa, Florida

People are not adequately prepared for disasters. 

Often, people are overconfident about their ability to deal with a disaster. They feel like they are prepared; however, in actuality, their plans are misinformed or inadequate. 

This effect is especially pronounced in people who live in hurricane-prone zones, since they feel like they have had plenty of experience. 

Unpreparedness is a problem, since it can lead to paralysis and poor decision-making when the disaster strikes. 

People seek comfort and familiarity during a disaster. 

Hurricanes can cause strong emotional reactions for people, especially those with special medical needs, who are more affected by sudden changes in their environment. For this reason, people seek comfort and familiarity, and will prefer to shelter in place or at a family or friend's house

However, this can be a problem, since family or friend's houses may not be able to provide the resources someone with special medical needs requires. 

People do not consider medical shelters as a viable option during a hurricane. 

Medical shelters are established by the county especially for citizens with special medical needs, and will provide them with essential resources such as electricity for their device. However, people are often not aware of them or are doubtful about how comfortable they will be. 

Many medical shelters also require pre-registration before the hurricane for planning purposes. Because people do not prepare, these shelters are often under-registered and may be forced to turn away citizens who show up at the last minute.  In one county in Florida, there were only 991 registrants out of 27,000 eligible. 

A hospital becomes a hurricane shelter of last resort. 

Rather than going to a medical shelter, people with medical needs will go to hospitals, since they know that a hospital is a comfortable place, where resources (such as electricity and medical staff) will be provided. Many people also see the hospital as part of the community, providing them with emotional comfort. 

hospital shelter.png

hospitals are overwhelmed by the surge of citizens seeking shelter

The surge of people seeking shelter can create unique challenges for hospitals, including:

We ended up with a thousand more people than we expected during Hurricane Irma.
— Hospital director from Florida
  • Tracking non-patients in their system.

  • Finding physical space to house non-patients.

  • Supplying enough resources, such as food, linens, and medical staff.

Although hospitals recognize their role in the community and want to help these vulnerable populations, they don't want to do this at the expense of patient care. Ideally, they hope that those without an immediate medical need will shelter at a different location, such as a medical shelter. 

Solution Goals

Taking into account the challenges faced by both citizens and hospitals, we crafted four specific solution goals. 

solution goals.jpg

Our Solution

Keeping our solution goals in mind, we designed Haven. 

Haven is a mobile app that redirects citizens with special medical needs away from the hospital, by helping them better prepare for hurricanes, as well as guiding them to an alternate shelter. 

The main intended user of this app is a caregiver, who is often the one in charge of making emergency plans and keeping the person with special medical needs safe. 

user types.png

How it Works

(GIFS created by Ketki Jadhav)

Log In

Customized flow based on user type, so that users are shown information and given tasks based on their specific and unique needs. 

Caregivers of citizens with special medical needs are the main intended users of this app, since they are more likely to be comfortable with technology than the patient. 


Medical Shelter Registration

After logging in, caregivers are immediately encouraged to complete the form for medical shelter registration. 

To make the experience less overwhelming and easier to navigate, the form is divided into cards by category. 

Progress bar on each card provides motivation to continue completing the form. 

Ability to share form and collaborate, since multiple people often take care of the same patient (ie: two sisters taking care of their mother). 




Medical Shelter Registration

Based on inputted information, the form will adjust so that only sections that are required for the specific patient will be displayed, ensuring that the experience is as quick and personalized as possible. For example, a COPD patient will be required to fill out the oxygen dependency card. 

Information icons provide additional explanation on medical terms, making the process simpler for the average citizen. 



Preparation tasks are chunked by time and category, so that the process is more manageable and will not be left to the last minute. (To ensure that our organization of tasks matches the users' mental models, card sorting was conducted with participants). 

Overall progress bar and picture at the top of the screen provides motivation to continue completing tasks. 

Based on inputted information, the app will only display sections that are relevant to the user, reducing cognitive load. 




Information icons provide justification for the app's recommendations. This transparency builds trust with the user and prevents skepticism from people who already feel like they know what to do. 



Shelter recommendations are prioritized and reprioritized based on the patient's  needs and preferences, as well as changing environmental conditions (such as road closures or power outages). 

Clear, authoritative directions reduce the user's cognitive load and prevent the paralysis that can occur during disasters.

Justification and reasoning behind recommendations provides transparency, builds trust, and gives users peace of mind. 

List of shelter resources helps mitigates the doubt that patients feel about quality of medical shelters. 



Map view provides turn by turn directions to shelter and adjusts according to changing environmental conditions, such as road closures, saving users precious time. 

Link to external source (government hurricane zone map) increases app's credibility and builds user's trust. 

Methods to Enact Behavioral Change

To get citizens to change their existing behavior regarding disaster preparedness and response, we consistently incorporated certain psychological principles in our design. 


Personalization: People are more likely to respond to content that has been customized for them, since it reduces information overload and provides them with feelings of control. 


Transparency: People are more likely to accept a recommendation if they understand the reasoning behind it.


Progress: People are more motivated to continue doing a task if they see that each step makes a difference. 


Authority: People tend to trust and respond better to clear, authoritative directions during a disaster, since it reduces their cognitive load and removes the burden of decision-making during a stressful time. 


UI Considerations

Because the primary users of Haven will be caregivers (34% of which are 65 years or older) as well as people with special medical needs (who may suffer from cognitive, visual, or motor impairments), we paid extra attention to certain UI elements, including:

Screen Shot 2018-09-01 at 12.15.44 PM.png

Value for Hospitals

A co-design activity we conducted to uncover the data needs of a TeleTracking client that unexpectedly acted as an emergency shelter during Hurricane Irma.

A co-design activity we conducted to uncover the data needs of a TeleTracking client that unexpectedly acted as an emergency shelter during Hurricane Irma.

Screen Shot 2018-08-28 at 5.40.59 PM.png

Redirection: In the short term, by encouraging citizens to remain safely at home or to go to a medical shelter if they can, there will be a reduction in the numbers of people who choose to evacuate to a hospital

Information & Prediction: For those who cannot or won’t go to a shelter, our app will provide descriptive information about their condition and their household and with this data, hospitals can more accurately predict the numbers and the needs of the people who may be sheltering with them. 

 Product Future

Onboarding and Usage Strategy

Next Steps


Our Process- Exploratory Phase


  1. Immerse ourselves in the domain and learn more about disasters.

  2. Identify current problems in the domain.

  3. Narrow down to the most affected populations.

  4. Validate the client’s prompt.

methods used

Secondary Research
To gain knowledge about the broad domain of disasters, we analyzed news articles, academic papers, existing surveys, etc., allowing us to identify which one of our research questions were already answered and which questions we needed to answer via primary research. 

Guerilla Research
Purpose: To better understand how the general population deals with and prepares for disasters. 

Purpose: To get the point of view of hospitals, disaster survivors, disaster experts (such as the Red Cross), and better understand their problems.
These initial interviews helped us focus on a more narrow population: citizens with special medical needs. Having narrowed down to this population, we then conducted additional interviews with citizens with special medical needs, caregivers, and medical shelters


Purpose: To reach a broader audience and identify trends about people's behavior and experiences during disasters. 

Stakeholder Mapping
To map out all the entities in the disaster domain, identify key relationships, and narrow down on breakdowns. 

Our Process- Generative Phase


  1. Begin ideating solutions for the problems uncovered in the exploratory phase.

  2. Generate a breadth of ideas/opportunities via design activities.

  3. Begin validating the user need and concept with stakeholders and users.

Methods Used

Design Activities
To generate a breadth of solutions, we engaged in a number of design activities, including How Might We, Visioning, Crazy 8s, and Reverse Assumptions. 

how might we.jpeg

Storyboarding & Speed Dating
After generating these ideas, we were able to narrow down to 6 potential solutions, which we then speed dated with TeleTracking employees, hospitals, medical shelters, citizens with special medical needs, and caregivers.


With this feedback on our main design ideas, we were able to consolidate to one design concept to pursue in the next phase, which took into account the needs and pain points of both citizens and hospitals.

Our Process- Iterative and Evaluative Phase

Screen Shot 2018-08-28 at 4.43.16 PM.png


  1. Begin building out the concept.

  2. Rapidly design, test, and iterate, evaluating both the user value and usability of the solution.

  3. Ensure that the solution is research-driven and effectively addresses the user and stakeholder’s needs and pain points.

Methods Used


User Testing Round #1: Validation of general concept, testing usefulness of specific features
Paper prototype with analogous population in Pittsburgh

User Testing Round #2: Testing usefulness of specific features.
Paper prototype with caregivers and general population onsite in Florida

User Testing Round #3: Testing usability and usefulness of entire app.
Remote, moderated user testing with caregivers in hurricane-prone areas via Lookback (think-aloud protocol)

After each round of user testing (around one week), we synthesized the data into actionable insights and designs recommendations to incorporate into the next iteration of our solution.