As app developers, we always have a keen interest in how apps impact us in our daily lives. The rollout of a Contact Tracing (CT) app could significantly improve the current COVID-19 situation. We want to look at the benefits, options and challenges of rolling out a Contact Tracing app to the public and the best option from our experience.
Contact tracing is the concept of recording when people come into contact. Contact Tracing using mobile phones assumes that if two phones come close together, that two people have, and that a phone is a good proxy for an individual. Digital contact tracing works either using Bluetooth, ultrasonic audio, or a combination. Solutions that don’t rely on a user’s location are more private.
There are existing efforts through universities and open source projects to implement contact tracing, both COVID-19 related and much older. Many of the current open-source projects focus on privacy and security as key aspects.
A digital contact tracing programme would allow for better use of the limited supply of test kits available and would provide safeguards to allow relaxing the isolation measures sooner.
Singapore has released the TraceTogether app which was developed in 8 weeks this year and published as open-source (OpenTrace)
Last Friday, Apple and Google announced a joint programme to provide a Bluetooth and Cryptography framework (to be released “in May”) to allow apps to perform Contract Tracing more reliably, and to provide a standard.
In May, the app store will allow apps that use this new technology, but only from vendors that are official health bodies/governments. A few months from now, Apple/Google will make this tech part of their OS and a way to let users opt into it without installing an app at all.
As long as a CT app does not collect any other personal data such as phone number, or address, or location and is decentralised, there is little to no privacy concerns. This will be challenging to communicate to the public, especially with Android asking for location permission just to use Bluetooth. If Android doesn’t change this, this will be a hard problem to overcome (“trust us, we don’t track your location - but we still ask for permission to….”).
Ideally, the source code for the app should be open-sourced. However, how do you assure people that the app they installed is the same as the source they can see?
Singapore only achieved 17%-20% adoption (1 million installs). Other experts quote anything from 60%, 80% or 100% as “minimum for success”. (80% according to a UK report)
What does success look like?
To achieve the 40%-80% adoption rate the onboarding process needs to be extremely well-thought-out to ensure users will finish the installation. The onboarding process needs to be supported by a creative communication plan to get the user to even download the app from the app stores. Successful installation does not mean adoption. The app needs to be in use. Ideally, the app will need to work while in the background. This is one of the biggest hurdles for “real” adoption on iOS. The Singapore TraceTogether relies that iOS users have the app running in the foreground. Anyone who has done app development knows that this will definitely not work.
Bluetooth is a specialist subject and implementation on phones is somehow opaque/varied. We have worked with Bluetooth tech for many years now and still find it a very challenging technology to work with. Bluetooth Low Energy (BLE) has made things more straightforward, but still, the way it works across the different smartphones is not consistent. It’s complex.
It’s crucial to have CT apps in the market as soon as possible. Short time-to-market is a problem that is being solved by may product developers who are using iterative and agile development and design sprints to validate features quickly.
However, testing can be a challenge as Bluetooth requires real testing with real devices. In our experience testing Bluetooth, device connectivity requires much more effort than traditional app development. Testing in a hurry can result in lower quality apps.
For the governments to be able to make better decisions, they may be looking for as much data as possible, and users are expected to give up more privacy.
Tech companies such as Apple and Google are claiming to protect users privacy and reduce data exposure. This is an opportunity for tech giants to be seen helping consumers and also solving big problems at the same time.
The two opposing gaols of governments and tech giants create a healthy tension that could lead to an overall better outcome for the end-user. (UK NHS is already clashing with Apple/Google.)
For the Digital Contact Tracing app to be effective 60% - 80% of mobile app users need to install the app and have the app running. So, the solution needs to be a combination of user experience and technology. The critical success factors will be easy onboarding, ensuring the user trusts the app and the app working when in the background, which is not possible on iOS at the current time.
The Apple /Google approach is the only option to ensure the app will be running in the background for iOS users, and this will increase the uptake and therefore, the effectiveness of a contact tracing app better than any of the existing approaches currently available.