Geospatial Revolution Mapping the Pandemic

Show video

(somber music) (suspenseful music) (bell ringing) - [Peter Coyote] 1854, London. In the wake of a cholera epidemic, Dr. John Snow placed cholera deaths on a grid to confirm the source of the outbreak as water from a single pump. Geography is central to fighting disease.

New Year's Eve 2019, 166 years later, a global infectious disease surveillance system detects an unidentified respiratory illness emerging in Central China. The threat was real. And the platform issued the warning nine days before the official WHO report. The system was built by an interdisciplinary team at BlueDot, a small startup out of Toronto; using artificial intelligence and machine learning to scour the internet for clues of emerging infectious disease. - When we're talking about the problem of infectious diseases, we need a diverse set of data. We can't just be tracking a microbe, we need to understand how people are moving across the planet, we need to understand in some situations climate conditions, animal populations, health systems.

And so when we think about where geospatial technologies are going, there's this massive increase in production of data that are all associated with a place and time. And geospatial technologies now allow us to wield those data to integrate them and to make sense from massive amounts of data to derive an actionable insight and to do this in real time. - [Peter Coyote] In an unequal world where official reports of outbreaks are not always timely, epidemiology's pressing need is to develop surveillance capacity, to sift through the layers of data that can predict an outbreak. - So we've been using machine learning and natural language processing to stitch together a global panoramic view of infectious disease threats that are appearing around the world.

And so our algorithms are scanning through vast amounts of online sources, open source data, in 65 different languages and looking for these early signals that maybe there's an unusual cluster of a disease happening. It might be a disease that we know of, or it might just be a syndrome in the way that we in fact picked up an early signal that there was an unusual cluster of pneumonia cases in a city called Wuhan. [Peter Coyote] These global tracking systems, mine, process and filter myriad information sources to essentially create a living breathing map of risk information. And to build it takes a host of experts ranging from public health to data science to veterinary medicine. - This is inherently a complex problem, and it requires a diverse set of skills and perspectives. - [Peter Coyote] This emerging field of digital epidemiology and disease surveillance is a mashup of computer science, statistics and big data mapping.

Algorithms are trained to understand disease terms and to process whether an alert is relevant. And while detection is the first concern, dispersion of the pathogen is the second. - [Kamran Khan] Human population, mobility and movements are critical to understanding how diseases spread.

So humans in essence have become the vectors that are transporting diseases inadvertently across the planet. So within the span of about a second or two, we had automatically now assessed how Wuhan was connected to the rest of the planet and which places were in the path. If this thing were to spread out further, where should we be looking for it next? - [Peter Coyote] The only requirement for the spread of COVID-19 is proximity and movement.

Transported undetected by their human hosts, the virus crossed borders worldwide. (suspenseful music) At Johns Hopkins University in Baltimore, a small team led by Dr. Lauren Gardner had been watching the outbreak with deepening concern. - [Lauren Gardner] I saw this as a really unique opportunity to start building out a dataset on an emerging infectious disease in real time. - Dr. Gardner and I, we had a very small group team meeting.

She suggested to me why don't you create a dashboard for the coronavirus because she knew I was from China and very anxious to know what's the situation in China and know my relatives, my friends and how they were doing. - [Lauren Gardner] My student Ensheng Dong and I were basically the two that started this and he's a total wiz with dashboard development. - After I talk with Dr. Gardner, I went back to the office. I spent eight to nine hours and came up with the first version of the dashboard.

- [Lauren Gardner] And so we built the dashboard that night and then the next day shared it publicly. - The dashboard went viral after that. - I don't know how many views the Hopkins dashboard got and continues get, but I'm sure it's off the charts.

- They are in fact, the creators of the most viral map-based application in the history of the world. - Four or five billion maps are generated off that website every day. Sometimes it was as high as 10 and 12 billion on a day. That meant that people were looking at it and re-looking at it and seeing the trends. - [Gardner] We massively underestimated the interest of the general public in following the progression of this outbreak.

[Nita Bharti] For a lot of people being able to see disease data in a spatial format meaning an interactive map was a really important way for them to understand what was happening and also for them to assess what their immediate risk was. - This was a huge day of awakening for the world because they began to see the power of geospatial visualization in their own lives. - [Peter Coyote] The Johns Hopkins dashboard quickly became the authoritative source and public health officials rushed to create their own. - It was within a week that we started to see similar dashboards coming out and Hong Kong had a dashboard and Thailand had a dashboard and Korea had a dashboard. - And I remember the first time I saw it thinking, I want that including the map and the visualization of all the countries and all the cases in the red dots, I want that on our website. - Anybody who felt comfortable going into GitHub, could go pull those data, could pull the actual tables and the numeric data and make their own maps or do their own analysis.

It was very transparent, it was updated regularly and it was easy to use. - [Este Geraghty] Never before have we really had this opportunity to watch real time data update. And that is the ideal for disease surveillance. - And this is why visualization of that work and the work that we've been doing on the ground is so critical because we need to know where things are happening so we can address them. We can focus our attention, we can accentuate our efforts and ultimately we can drive resources to those very communities that are seeing the disproportionate impact of this horrific pandemic. (suspenseful music) - [Peter Coyote] Days before the WHO called the crisis a pandemic, the Johns Hopkins map showed it to be true.

The small clusters of red dots had grown larger and crossed borders. But dark spaces on the map marked where public health data had been withheld. One of these was Iran. In late February 2020, Iran's Ministry of Health reported 12 deaths, but posts on social media painted a darker picture. A team at Maxar Technologies used their human geography dataset and crowdsourcing platform to investigate the rumors, then confirmed them with high resolution satellite imagery. - The largest cemetery in Iran in Qom, what we were seeing from space... mass graves being dug.

And so we saw trenches that were upwards of a hundred yards long. We saw large piles of lime to counter the odor and the decay. - [Peter Coyote] The satellite pictures suggested the number of dead was significantly higher than the official figure.

- That then allowed us to be able to support in this case The Washington Post, to be able to publish an article to bring transparency. And I think having that type of shared experience where we all were able to see the impact of that, I think it helps drive behavior. (lively instrumental music) The goal is to understand what is normal and when is there a variance. This proliferation of sensors, whether those sensors are on the ground, through social media or passive sensors that are collecting data.

If we can derive patterns from that, that then can help inform how we apply assets in space to be able to collect imagery and then extract the insights from that imagery in short timelines to be able to drive decision-making. - [Peter Coyote] From space, the signs of COVID could be seen everywhere. Passenger planes grounded, new hospitals built within months, St. Peter's Square on Palm Sunday empty.

In Mecca, the crowds that visit the Kabaa, Islam's holiest shrine, told the story. - Going from densely populated cultural iconic location of interest to no one there and then the leaders at that location established mechanisms to create social distancing kind of a stark change in behavior. (suspenseful music) - [Peter Coyote] On February 29th, 2020, the Trump administration announced new travel restrictions on Iran. The same day the US recorded its first death from COVID-19. Colleges canceled student travel programs, but many students had already boarded planes for spring break.

- You see on news that spring breakers are still flocking to the beach in Fort Lauderdale. We thought, man, this is really reckless to have all of these people go from universities to one location, congregate who don't really feel the symptoms of COVID get together and then return back home and go back to their schools. We said, let's see if we can show that story in our mapping platform". - [Peter Coyote] To bring awareness to the situation, Tectonix, a data analytics and visualization company, planned to isolate and track cell phone users crowding onto a single beach using anonymized location data. - [Rob Gresham] As we zoom in towards the Fort Lauderdale Beach, we are saying "filter the geo-fence down to what you're actually seeing on the map". And then what you end up seeing next is an even more precise filter where we start drawing a lasso around devices on the beach.

And you could just see the congregation of many many people. - [Peter Coyote] Once lassoed by the geo-fence, every beachgoer's cell phone location data could be tracked by Tectonix platform. [Rob Gresham] And then when we zoomed out in the days following, you could see these students returning to essentially all over the country.

- [Peter Coyote] The team sent out a video on social media. - The next morning, checked Twitter and saw that we had millions of views. And we just said, "what the heck have we gotten ourselves into?" - [Peter Coyote] On March 15th all public beaches in Fort Lauderdale were closed. (lively instrumental music) By August, US COVID deaths had reached 160,000. Warnings about mass gatherings and social distancing largely went unheeded. And in South Dakota, the 2020 Sturgis Motorcycle Rally opened as scheduled, 250,000 attended.

The Tectonix crew lassoed devices in all the hotspots and then tracked them back to their home states. - Sturgis was interesting because it happened to be a lot of middle-aged or older people. But when people flocked to South Dakota, it actually set off a series of events where South Dakota briefly was the epicenter for COVID spread globally for several weeks. - [Peter Coyote] In rural America, meat packing plants reached a crisis situation with COVID cases.

Tectonix isolated an Indiana plant for a full month, the resulting footprint was huge. Devices traveled to 48 states and into Canada. - For whatever reason, the meat packing industry was very hard hit by COVID. And now I think we've seen a lot of the working conditions and how close people were on conveyor belts and that sort of thing.

- [Peter Coyote] These demonstrations proved in case after case, how desperately interventions were needed. That responsibility fell to public health officials who found themselves facing an angry backlash against emergency measures as they prepared to roll out the largest vaccination campaign in American history. (suspenseful music) - Geospatial technology has been a huge asset for epidemiology... for understanding

how outbreaks are moving through populations and understanding how we can get response to those populations. - Now you've got all this data, how are you then determining what you do there? How do you accentuate your communication messages? What are you pulling back? If it's a zero sum game you don't have enough resources, you can only do one or the other; you're gonna focus on where you have the most risk. The most proportion of tests that are coming back positive, the most hospitalizations, what's happening with deaths in that area.

But then it's the next phase of it which is now about adding in the protective measures. Where are your PPEs? Where are the masks going? What are the data showing in terms of mobility? Are people wearing masks? Are they moving around? Are they not moving around? Are they staying close to their home? This is all geospatial technology and mapping that allows you to visualize. And then on top of that the last protective measure is on vaccines. - To monitor our vaccine inventory, to understand the population phases and how many people in each phase need to be vaccinated.

To look at the vaccination venues. - Now you layer socioeconomic information, housing, transportation issues, challenges with social vulnerability index or what we call them SVI. Add in where you have increase in cases, hospitalizations, deaths and vaccine knowledge. And now you have a geospatial ecosystem that you can not just share internally, but share with your folks on the ground but also your policy makers, your elected officials, with the community who are looking at all this information and saying, "Gosh, I wanna protect my own family!" (suspenseful music) - [Peter Coyote] As vaccinations take hold across the United States, COVID-19 cases and deaths have dropped to their lowest levels in nearly a year. But even as the outlook improves, researchers are testing new methods to stop the next pandemic. A team at Scripps Research in San Diego has proposed using wearable technology to monitor a population for early signs of an outbreak.

(heart beating) - So we're launching DETECT and we're really excited about this. Because this is a way to track from a smartwatch, a person's heart rate and to know well before people have fever and symptoms, whether there's a cluster of abnormalities that's happening. - [Peter Coyote] Roughly a hundred million Americans own a wearable health tracker.

And this study only needs to gather physiological data from one or two percent of them to set up an effective public health surveillance system. [Eric Topol] We have a mobile app where people will be asked if they wanna join a research effort to track their health, their resting heart rate through their smartwatch. - [Peter Coyote] Once a person volunteers, readings from their device will be monitored continuously through the app. And they'll be asked to enter their vaccination dates and report symptoms. - [Eric Topol] In the setting of this COVID-19, it gives us a handle on things like heart rate, which we know is a great antecedent abnormality before people get sick. So by tracking people at scale, we'll be able to see an outbreak in a way that previously has not been obtainable.

- [Peter Coyote] This new twist on telemedicine, the ability to perform medical interactions without face-to-face contact is especially important in the midst of a pandemic; when we simply can't test enough people in time to know what the virus is doing. - It's real time, it's continuous, it's scalable. And so we're really keen on this added ability to be able to detect an emerging problem in any part of the geography of the United States. (somber music) [Peter Coyote] The global pandemic has caused incredible pain and hardship.

We have suffered great loss and isolation, and we have not seen the last of COVID-19. The reality is that the world has entered a pandemic era in which new pathogens continue to emerge and spread within populations. And more than ever, geography is central to fighting disease. - [Kamran Khan] COVID-19 I think has been an awakening and we can all agree none of us wanna be back here again soon.

- We certainly have hope with the vaccine, but concern yet about the variants. We have a number of lessons learned, but it's unknown how well we'll do with applying those lessons. - I'm trying to think about it as an opportunity to make changes and to think about the world in different ways. What can we be doing? What kind of information can we be collecting? The limit to this technology and the limit to geography is really our imagination. - COVID-19... what it has done,

it's allowed us to see a purpose that's bigger than ourselves and I think deep down, probably every one of us wants to feel like we're part of something that's just bigger than us. - Geospatial community stood up and I'm so proud of them. Their work has literally saved lives and made things understandable. It's a great day for the application of science and technology to this great challenge that's facing the whole world. (light instrumental music changing to somber music)

2021-07-30

Show video