The federal government believes that social distancing restrictions can be eased sooner if there is a greater surveillance of people infected by the coronavirus and much faster contact tracing for stopping the spread. In fact, we can learn from how the South Koreans are managing the spread without a large-scale lockdown by rapid testing, tracing and containing those that are infected while using technology. Hence, it is very encouraging to learn that after the covidsafe app was released, more than a million Australians downloaded the app, reaching the five-day download goal within five hours. However, this early indication cannot be a measure of success, or even a reason to celebrate just yet.
Without public trust, the covidsafe app cannot succeed”. Why does trust matters so much that it is the key factor for success? Without getting too technical, it is useful to learn about how the app works to help the medical workers and the state government trace and be one-step-ahead to stop the spread.
Illustration of how trace and contain – Source: PEPP-PT
In a nutshell, the app records two people’s phone details when they are within proximity for a period of time long enough to risk a spread. The encrypted record will be stored locally at both phones with the time-stamp. No location data is tracked or kept. When a person is diagnosed positive with the coronavirus by the local medical authority, that person will be prompted and assisted by the health official to upload the data from his/her phone to the cloud using a PIN. Then, the data can be decrypted by the official to identify and alert those that have been in close proximity to consider testing (summarised from the app’s official website, refer to it for further details).
The government is releasing the source code of the app, so that public can be assured of what goes on under the hood, and be assured while using the app. Moreover, the government is also considering legislating a new law for regulating the intention of use, so that Australians can have absolute certainty that this data will only be used for the very specific purpose: protecting citizens and health purposes.
Even if the app is downloaded by all Australians, the remaining question is: would it actually support the intended use effectively? Learning from Singaporean government’s strategy to have every citizen using the app, the biggest caveat is that they are still in the very early stages of using the app, hence it is too early to tell how effective the Singaporean app (Tracetogether) actually is.
This is in contrast with the Australian PM’s optimistic target for minimum required uptake, which is just 40% of the population, to make the app effective. Given that it takes two phones to register a contact tracing record, if the ratio of users is too low, the traceable record will be incomplete and ineffective if there is not enough phone with the app. Moreover, after download (app installation), all users need to use the app to capture accurate (and not fake) information, and keep their phone running at all times. Otherwise, the gaps in data would still need to be filled in by manual “detective” work by nurses, doctors, and public health officers for retracing patient’s movement. Therefore, the key factors of the app’s success are: ratio of active app in-use, correct/valid data entered on each app, and long-term regular use of the app. These are the key factors that are far beyond the celebrated and promoted measure of uptake.
To help improving trust for increasing uptake and long-term use of the app, the government has done everything they can to give assurances for Australians to trust the usefulness of the app. However, there is a real risk that this app may not become the silver bullet that was promised and hyped to help Australians to maintain the spread of coronavirus. A statement from the chief medical officer that “no Australian should have any concerns about downloading this app” (ABC), is optimistic at best, and could in fact backfire if there is a single episode of the app’s privacy and security breached. For a start, some experts already warned there is still a risk that the data can be obtained by foreign entities and government (ABC), which would be a serious breach of privacy in spite of all the promises by our government that the data will be appropriately managed to used. Hence, it can be anticipated that early adopters of the app may kept on installing and uninstalling the app as they battle through the trust issue. Given that the contact tracing data is only kept locally on the phone and only uploaded when required, the app must stay on the phone to ensure local storage persistence. If the app is not constantly kept and used regularly on the phone, not only that the data is no longer reliable enough, but also much of the valuable data would be lost as deleting the app will also remove the data from storage (covidsafe).
In addition to mobile app, the government could consider more localised solutions to enforce social distancing in public spaces. A video presentation of our latest research output ready for deployment: “Privacy preserving video technology for social distancing” is a spin-off from a recently published paper that won a best runner-up paper in an international conference in the Advanced Video and Signal based Surveillance Conference 2019. To maintain social distancing during covid-19 pandemic, it is important to prevent crowding in public spaces. Our state-of-the-art technology uses real-time embedded computing to automatically count the number of people and monitor their spatial distancing in surveillance cameras without storing the actual image information for maintaining privacy. Hence, authorities can be immediately alerted when crowd density is becoming critical. The biggest caveat with this technology is: video surveillance has always been seen as the big brother approach, and Australians generally are not comfortable with having their images being stored. So, it remains a question whether such technology to preserve privacy by removing image data would be more acceptable.
Illustration of our privacy preserving video technology: the embedded computing in the camera will only collect and use the information on the right (non-image – only spatial distancing between detected people) to help local authorities enforce social distancing.
So, what does it take for technology to support Australia returning to normal? The answer lies on the initial trust formation, and the continuous trust development through a very transparent process of aligning the technology with the user’s expectations and comfort zone, while ensuring that the technology is consistently meeting expectations and compliant to the regulations and basic human rights of privacy and freedom. No technology can be a silver bullet, hence there isn’t one that can be prescribed to solve the goal of returning to normal amid this current pandemic. Instead, we should continue to explore and test a combination of technology to support a common goal, which is minimising spread of infection without a large-scale lockdown, which is hurting both the economy and people’s long-term mental health and wellbeing.
Professor Dian Tjondronegoro is the Deputy Head (Research) of the Department of Business Strategy and Innovation, Griffith University.