7.13.2020 | DMG Blog |
As COVID-19 advances worldwide, social distancing and contact tracing have become key elements in our fight to contain the disease. Knowing who has come in contact with an individual who is COVID-19 positive helps citizens, health organizations, NGOs and governments analyze the spread of COVID-19 in a given area and alert someone who has come in contact with a COVID-19 positive individual.
Since the start of the outbreak, the term “social distancing” has found its way into the vernacular. COVID-19 can be transmitted through close proximity to affected individuals and the established guideline for “close proximity” is typically 6-feet (2 meters). Public health officials have identified contact tracing, or the analysis of who has come in contact with an infected person, as a valuable tool to help contain the spread of the virus.
However last week, the World Health Organization (WHO) agreed to formally acknowledge the threat aerosols play in the transmission of the virus in indoor spaces. This move was prompted after 239 researchers from 32 countries sent an open letter to the WHO. They argued that microscopic particles are released during exhalation that remain in the air and travel further than the established 6-foot social distancing guidelines set for indoor spaces like markets, restaurants, bars, offices and other businesses.
Regardless of the debate over “proximity”, the fact remains that contact in indoor settings can promote the spread of the virus. Therefore, many leading public health authorities, universities, and NGOs around the world have developed opt-in contact tracing technologies. Most of these applications, including the much-heralded collaboration between Apple and Google, originally planned to use Bluetooth technology to detect when one device is near another, and for how long.
The Apple Google COVID-19 Ecosystem
The first phase of the system lets health agencies build apps that allow a person who tests positive for COVID-19 to input their diagnosis. The system then uses Bluetooth technology to learn who the person has come in contact with and then notify those people of a possible exposure. Ireland is the latest European country to successfully launch a national contact-tracing app that utilizes the Apple Google technology to track down and warn people who have been in contact with someone who has tested positive for COVID-19. The app, called COVID Tracker, leverages the Exposure Notification API developed by Apple and Google.
For the app to work effectively, users must enable Bluetooth. The phone will then continuously scan for other phones nearby that have also downloaded the app and have activated Bluetooth. COVID Tracker then anonymously logs the distance between devices and the length of exposure.
Every two hours, COVID Tracker will download a list of devices that are tied to users who have tested positive for COVID-19. If the user has been closer than six fee (2-meters), and for more than 15 minutes, with any of the phones associated with the positive tests, they will receive a notification that they are a close contact.
This All Sounds Great. So What’s the Problem?
It is our position that mobile apps will fail to deliver useful contact tracing results for three reasons.
The first problem with using mobile apps for contact tracing is these apps may not be accessible to enough people to have a worthwhile effect. Let’s assume that for contact tracing to be effective, you need at least 75% or more of the population using it. After all, if only a small number of people are using it, it will be difficult to trace enough people to reasonably determine if someone has come in contact with a positive case.
But many of the apps that support the Apple Google API are only usable on newer smartphones and more importantly, the use of smartphones is limited among older generations who are most vulnerable. My father-in-law is classic example. He refuses to purchase a smart phone and remains content with his early 2000 flip phone.
Furthermore, China has banned Google’s technology. So they will need to adopt an alternative system that does not communicate with the Apple Google led effort.
At least 2.5 billion people will not have the option of using a COVID-19 app tied to the rest of the world. This includes people who do not have a smart phone or who have older smartphone technology, along with the population of China. Therefore, getting a 75% or higher rate of adoption is a large hill to climb.
The second problem is they are optional. There is no enforcement or requirement to use these apps. Users not only must volutarily download the app and voluntarily provide the required input to set-up their profile, including the COVID19 symptoms screening questions. They also must enable Bluetooth and Location Services. Assuming that a large portion of population does this right and voluntarily, further diminishes the adoption rate.
The third problem has to do with the efforts being made to protect user privacy. In an effort to preserve privacy, the location of people is not tracked (or exposed) as the location is only stored on the user’s phone. Since there is no central repository to mine, health experts are unable to collect information to identify concentrated outbreaks.
And that leads us to the real issue in the Google/Apple take-over of the whole smartphone contact-tracing ecosystem—it has put privacy before efficacy. Let’s be very clear, you can prioritize privacy or you can prioritize fighting the spread of the infection, [but] you can’t have it both ways, you can’t have your cake and eat it—to steal that unfortunate U.K. analogy.Zak Doffman, Contributor at Forbes and Founder/CEO of Digital Barriers
AI-Driven Solutions Look to Video Surveillance to Mitigate User Adoption While Maintaining Privacy
We believe a viable alternative to smart phones and contact tracing applications is the use of video cameras and AI technology to analyze the movement of people in a given area. In an effort to promote social distancing, Amazon recently announced Distance Assistant, to provide social distancing behavior in real-time. The tool helps associates see their physical distancing from others using green and red circles around each person camera’s view. Distance Assistant provides employees with live feedback on social distancing via a 50 inch monitor, a camera, and a local computing device. The standalone unit uses machine learning models to differentiate people from their surroundings.
At DMG, we are developing a similar system, using motion-based multiple objects tracking, called DMGsmart Track. The purpose of DMGsmart Track is to perform contact tracing in places like theme parks, sports arenas, music venues, hotels or cruise ships. It is especially useful in places where entrance to the venue is tracked by some form of identification.
As the name implies, DMGsmart Track tracks the movement of people using one or more cameras and determines each time that two or more people come within 2 meters (6-feet) of each other. This is called a “contact incident”. Each contact incident is stored in a database where they can be “traced” back later to determine who was in close contact with someone who may have tested positive for COVID-19.
DMGsmart Track uses existing digital close circuit camera technology with facial and image recognition to identify people in a given area. Each person is labeled in the application with a unique identifier that does not include any personally identifiable information (e.g. name, DOB, SSN). When someone tests positive for COVID-19, their image can be matched up with the system of record responsible for guest management.
We believe that AI-driven technologies using digital cameras can help reduce the challenges posed by mobile device accessibility and voluntary use of the COVID tracking apps, while maintaining privacy.