Overview

Problem

As we began to relax quarantine restrictions, we needed to find a way to more efficiently monitor transmission to prevent additional waves of virus transmission. At that time, isolation and quarantine measures in place were in the hope of reducing occurrence of novel cases and flatten the curve, but we needed something to expedite and aid the process.


Solution

Contact tracing is an integral practice that the healthcare system employs to ensure isolation of at-risk contacts and curb further spread but it is all done manually; so we created an app that did all of that! CanTrace bridges technology with healthcare to create an effective tool against the infection, expedite recovery, and preserve civil liberties for the common good of all Canadians.


Timeline

16 weeks


Team

~34 (medical professionals, public health specialists, privacy experts, developers, me!)


Roles and Responsibilities

I was the sole product designer on the team. I worked through the entire design process with the team to create the mobile app.

We're in this together

DISCOVER

Understanding Contact Tracing and our Goal

Contact tracing is a process involving a public health agent identifying, assessing, and monitoring individuals infected by a communicable disease and their exposed contacts to mitigate transmission. While contact-tracing is an effective tool, it is labor-intensive, time-consuming, and subject to recall error. These shortcomings became strikingly evident due to COVID-19’s marked infectivity, particularly in densely populated environments, straining public health resources. As a result, our goal was to work on a digital contact tracing solution that involved using smartphone apps to track proximity interactions between people, which addressed the limitations of physical contact tracing due to their significant speed and scalability potential. We wanted to create a product for Canada as contact tracing apps natively help expedite and centralize the entire process to make it simpler.

DISCOVER

Secondary Market Research

In order to understand our audience base, identify new opportunities and set realistic targets based on concrete data, we conducted market research in two key areas. We looked into the rate of downloads of contact tracing apps in many countries to set realistic targets and investigate why they are so high/low. In addition, we researched Canadian government survey data to understand our target market and who exactly we were building our product for.

Willingness of Canadians to use a contact tracing app
COVID-19 contact tracing app adoption rate in different countries
Takeaways:

1. Age is an important factor in the likelihood of using such an app as older Canadians are more likely to use the app than young Canadians.

2. Concerns about privacy and government access to location data are the leading reasons against using a contact tracing app.

3. Secondary reasons are poor healthcare integration and guidance in-app.

4. The above factors seemed to be leading factors in downloading contact tracing apps in other countries as well.

DISCOVER

Figuring out the Right Technology

Keeping the user in mind, we investigated several different technologies to conduct tracing safely and reliably as that was a key factor for users. Various technical modalities exist to determine proximity traces, including Bluetooth Low Energy (BLE), Global Positioning System (GPS), WiFi and Quick Response (QR) codes among others. However, GPS and BLE dominated global discussions and were the most commonly used methods. GPS configures proximity by using timing and location measurements to identify overlap between app users’ paths. It poses the benefits of providing location context of exposures, and critical geolocation data to public health to identify infection hotspots and analyze disease spread. However, GPS is unable to resolve between objects within 4.9 meters of each other, and signal accuracy is undermined near physical barriers (like tall buildings).


BLE possesses key advantages over GPS including its ability to detect short distances between individuals and better discern proximity between those confined to the same geographical coordinates yet separated by elevation like in multi-storey buildings. Additionally, it is operational underground such as in tunnels or subways where satellite signalling is obstructed. Importantly, BLE’s capacity to configure proximity traces without processing or storing sensitive location data draws fewer privacy concerns. Based on these factors, we focused on building our product around BLE.

DISCOVER

The Apple/Google API

Contact tracing apps using BLE were initially hampered by a limitation on devices using Apple’s iOS operating system which required phones to be perpetually “awake”, and apps open in the foreground to keep BLE readings active. Furthermore, BLE communications were stifled from signalling inter-compatibility issues between Android and iOS devices. Consequently, app adoption rates were significantly reduced, which further minimized their contact tracing potential. As a result, Google and Apple partnered to create an Application Program Interface (API), a program that mediated interactions between different software, which effectively resolved these limitations. The API was also lauded for preserving user privacy by operating within the decentralized framework and barring the co-existence of GPS technology on contact tracing apps adopting it. Consequently, due to its technical and privacy advantages, it became the standard operating software for contact tracing apps like ours.

DISCOVER

Competitive Analysis

After figuring out what technology and how effective certain methods can be, we collected data from a few different contact tracing apps and investigated what may be the reason behind its low adoption rates, technologies they used and how much they kept the user's privacy in mind. After conducting this research, it only supported our prior research and we knew we were on the right track to solve the user's biggest problems against such an app.

Takeaways on low adoption rates:

1. Bluetooth tracking technology using BLE is safer than GPS tracking.

2. Integrate Apple/Google API to optimize performance and security.

3. Very shallow level of healthcare integration and user guidance.

DISCOVER

Storyboarding

We mapped out our user's journey to understand what kinds of interactions our product needs to support. And outline the steps the user needs to go through to successfully use our product.

Robert would then receive a notification on CanTrace and be guided by healthcare professionals on how he should take care of himself directly. This is how we were able to empathize and understand the real impact of an app like this. Our oldest are the most vulnerable against this virus, so let's do it for them and everyone on the front-line.

DISCOVER

Understanding the User

Taking what we learned from our research, storyboarding and our problems in mind, it was important to understand the types of users who were going to be using the product. Going back to these personas throughout the process kept us aligned with our user goals, empathize and helped us solve the problems more efficiently. In addition to our research, knowing that this virus does not discriminate and has no barriers, we knew the most important people we were creating this for were the elderly. And on the other side of the spectrum, we had someone younger. What this also helped us understand better is how simple and easy-to-use we must create our product.

DEFINE

Problem Definition

Through our research, we narrowed it down to three main points to define our problem for our product we were creating. Already released contact tracing apps fell short, they only focused primarily on notifying users if they had an exposure risk and nothing more.

DEFINE

User Flow

What's missing from current apps?

DEFINE

Our Solution

Our design recommendations operated within the API framework, capitalizing on its privacy and technical advantages, while incorporating user-centric features and healthcare elements to integrate conventional contact tracing efforts. Broadly, these recommendations augmented the API exposure notification platform with evidence-based characteristics to optimize both privacy and efficacy, help maximize app adoption rates, and minimize disease transmission. Without healthcare integration, apps did little to support contact tracing efforts by public health. Our solution was the bridge between technology and healthcare to create an effective tool against the infection, expedite recovery, and preserve civil liberties for the common good of all Canadians.

DEFINE

Revised User Flow

DEVELOP

Wireframes

These early wireframes provided an overall idea of how we wanted the app to look. They provided a clear vision of how our users are going to interact with our product and allowed us to further investigate any pain points.

DEVELOP

Iterations

Implemented app branding into visual design striving for a more professional look. By the third set of screens, we were right at the final design stage receiving direct feedback. We knew exactly how users were going to interact with our app and the overall design to follow.

DEVELOP

Measuring Success (User Testing)

There are many ways of defining KPIs to measure the success of our implementation and design. Using the prototype I built, we conducted user testing to ensure all areas of the app were implemented correctly and users were able to complete various tasks seamlessly. What we learned from these was that we created an app that was extremely easy to use and straightforward - considering people of all ages were going to use it.

Tasks:

1. Onboarding, set up local health unit

2. Uploading positive diagnosis

3. Completing self-assessment after close contact

Completion Rate

12/12

Error-Free Rate

11/12

DEVELOP

Design System

DELIVER

The simplest and seamless onboarding process out there

The idea of getting your users familiar with the app and its features is extremely important. This process allows us to strategically communicate with our users and facilitate a positive user experience right from the get-go.

You're also stratified to your local public health unit

To enable accurate healthcare integration while maximizing privacy and security - the user's province and postal code is solicited to assign them a unit closest to them. No personal information is saved.

DELIVER

If you've tested positive for COVID-19

Just enter the verification code texted to you by your public health unit. The app will then provide you with the resources and material to help guide you through your diagnosis.

DELIVER

If you came in close contact with someone who tested positive for COVID-19

Risk levels may vary on the situation. Therefore, we implemented a contact tracing triage self-assessment tool to assess your risk; and provide advice accordingly.

Learnings, Reflections and Outcome

When you're building a product that can potentially be used by 35 million people, the way your mind thinks changes a little. You get that extra motivation to complete the product with pristine perfection and work with your colleagues in going through the entire design process thoroughly and effectively. The most challenging aspect of the whole experience was to earn the trust of the user. Privacy and security were the biggest reasons people had against using such an app. So as a product designer, it was key to empathize with the user and understand that you need to design an experience that is simple to use and earns the trust of the users directly. All while keeping in mind that this app would be used by people of all ages. The way I solved this was looking at other apps that have a large age demographic and have a privacy aspect to it. In addition, I frequently communicated with our lawyer and privacy experts about the design and content to perfect the design.

By the time we presented our product to the government, they had already started developing their own app with Shopify and Blackberry and so we did not have much of a chance. But in retrospect, some features of our app were added into the COVID Alert app officially released by the federal government. The biggest learning I would take from this whole experience is to think swiftly, you never know when the opportunity may close.