Mass surveillance won't stop corona. What can tech do though?

Wouldn’t it be great if governments could trace the whereabouts of every person with SARS-CoV-2? That way they could call for the people who shared locations with those confirmed cases to go into quarantine. Since a lot of location data is already gathered by various digital systems, this shouldn’t be to difficult.

I’ve seen such proposals a couple of times popping up in news articles, but does that make sense? I think mass surveillance would be impractical and ineffective to fight the pandemic. I’d like to propose a better solutions, based on how people are motivated and how they behave.

But first, let’s briefly go into likely effects of putting a mass surveillance in place. Effect that we’d experience for decades—if not centuries—after COVID-19 vaccins are available.

Technology doesn’t care about intentions

Once a surveillance system has been put in place, it can and will be used for other purposes. If you can use it to track COVID-19 case, why not for other diseases too? If you can use it to trace patients, why not terrorists? If it works with terrorists, why not for murderers? If it works with them, why not with motorists behaving dangerously? Why not use it to make people behave better at work? And at home?

A surveillance system would start with location data, but we’d want biometric and health data in there as well, like heart rate and body temperature. Combined with image recognition techniques, the system could infer how excited, interested or put off people are. Very useful stuff for rulers of totalitarian states. Very dangerous for democracies.

Mass surveillance systems are sensitive to corruption

It won’t be doctors who will operate the surveillance system. More likely services like those hired by Facebook to do content monitoring would do checks and contacting of potentially infected people. In other words: the system would rely on big numbers of low-paid remote workers with little job security getting access to a population’s most sensitive personal data. That would be prone to corruption. Want to get some dirt on a political rivals, blackmail judges, intimidate activists? Surely a cheap minute-by-minute report of their whereabout helps.

The money should be spent more effectively

Big technology projects are expensive. The HealthCare.gov website alone cost over $2.1 billion to build. The scope of that project is tiny compared to a mass surveillance system involving all citizens using it every minute of every day.

As is so painfully obvious, such money would be better spent on healthcare systems instead. Let’s use that money to pay nurses and other hospital workers more decent salaries.

Digital mass surveillance can’t stop the pandemic

There may be people who don’t care about those long-term consequences. People fearing for their health may feel that something must be done right now, no matter the costs.

Unfortunately for them, putting mass surveillance systems in place just wouldn’t work.

Tracking technology is inaccurate

The location services on mobile phones combine GPS and information from cell towers, nearby wifi networks and bluetooth devices. This can be a lot more accurate than cell tower triangulation alone, but inside and near buildings, it’s still pretty rough.

Have you ever looked at the map in the photos app on your phone? It may show you where you’ve taken the photos in your library. Most likely, you’ll find pins scattered around your home. Pins for photos that were made inside your home. When it comes to tracking inside what buildings you’ve been, location services aren’t very useful.

Now let’s assume we’d have the technology that makes those pin clouds around buildings accurate. You’d then be able to see at what time you’ve been in what building. But that’s still not very useful. Several hundreds of people may live or work in a single building. I lived in a 14 story apartment flat and I’m pretty sure even after half a year I haven’t come near to the vast majority of residents. If I’d been tested positive for COVID-19, putting the whole building in quarantine would be overkill.

In China’s mass surveillance system, location data is combined with facial recognition. The exact locations of the cameras are known by the system, but facial recognition is still error-prone and can only be applied at a limited number of locations.

Surely technology improves. But right now, we care most about indoor positioning (where contagion is more likely) and that’s where location services perform the worst.

Building mass surveillance systems takes too long

Countries that value personal freedom don’t have dragnet surveillance systems like those exposed by Edward Snowden. So even if we’d have perfectly accurate location data collected by our mobile devices, there’s no centralized access to that information. And even with a centralized eavesdropping system like the NSA’s can’t be used for our case tracking purpose without major investments.

Technology projects involving millions of people are not just costly, they take long. By the time they’d be finished, a vaccine and perhaps even a cure would have been developed.

Moreover, with the current number of COVID-19 cases, in most areas it’s too late for containment already.

Not everyone has a mobile and neither should you always carry one

Surveillance relying on mobile phones is not just impractical because the location data is inaccurate. Not all people have phones that can collect location data. Even here in Germany it’s not uncommon for people to have feature phones.

Using location data drains the battery—my ownphone lasts about 1,5 hours when I track a bike ride.

Especially with the news being distressing and misinformation being spread on Facebook, it would be good if people leave their phones at the charger a bit longer and go for walks without devices. Hashtag KeepYourDistance.

How can we create a system to contain the next pandemic?

For most countries, it’s too late for containment of the SARS-CoV-2 virus. Too many people are infected already and it’s unlikely borders stay closed indefinitely. But we can be quite certain that a new virus will show up at some stage. After all, we’ve seen SARS and MERS in just the last two decades.

I imagine a system could be developed that citizens use to track their whereabouts. Then when a participant is tested positive for a new, life threatening virus, they can submit where they have been during the incubation period. Other users can then get informed if they have been close to the patient.

So how could we build such an infection tracing system, without all the drawbacks of a mass surveillance system?

Start early, start small

Instead of creating a national system immediately, start with a small population. A commonly understood principle in software development is that scaling up too early slows projects down and makes them expensive because of the complexity involved.

Another benefit of starting small is that the system can be tested early. Usability, user acceptance and technology can be improved iteratively, while gradually rolling out the system to a larger population.

Make it voluntary

Hundreds of millions of people switched from working in an office to working at home during the last few weeks. Where some bosses allegedly feared their minions would just lay on the couch watching movies, it turned out workers remain at least as motivated to perform well. Many others discover that they’re actually more productive working from home.

In this light, it seems backward that some are asking for mass surveillance. After all, we don’t need an all-seeing overlord for what we are intrinsically motivated to do. Where the message of mass surveillance says “We watch over your because we don’t trust you”, a voluntary system says “We have trust in you and your support of the community”.

A system to which people willingly provide information, likely gets more and better information.

Respect users and their privacy

To make it attractive for citizens to join a voluntary monitoring service, their location data should be stored on their device only, inaccessible to others. That is, until a user is tested positive: then they can submit their location data to the cloud. The submission would be anonymous. Data older than the longest incubation time for viruses known are discarded automatically. This should avoid the risk that the data is used for other purposes by governments, police or commercial parties.

The location tracking shouldn’t drain device battery, so there should be an option for users to manually check in to places like offices, restaurants. Time for another Foursquare pivot?

Make it precise

Even accurate GPS positions aren’t useful enough right now. If an infected user visits the restaurant of a museum, but didn’t enter the exhibitions, that would make a big difference for whom should be warned about this. So besides raw location info, users should be able to add notes about what areas of a building they’ve entered.

Make it open and international

The spread of COVID-19 shows that international tracking of cases can be essential. So, although a monitoring system should start small, its effectiveness shouldn’t stop at national borders. A free, open source data model would enable governments and other organizations from different areas to use the same information.

These are first ideas. Some of them inspired by the (sidenote: This website from the Dutch government was quickly built to do symptom surveys of the population. Many signed up and submitted data, until the project was shut down when a security issue was discovered. ) project from the Dutch government. I’m sure there you can come up with some requirements for a democracy and people-friendly monitoring system. Let’s make our societies come out of this crisis more resilient and better than ever!