(CNN)If the year 2020 is good for anything, it's the lesson that during a crisis, anyone who builds a better mousetrap will find the world beating a path to his door.
A humble team at Johns Hopkins University in Maryland reminded the world of late poet Ralph Waldo Emerson's phrase when they created a real time tracking map of coronavirus cases and deaths.
And the world came to their door. They report that the site, plus downloads of its data, hosts three to five billion interactions daily. By their measurement, interactions include uses of the public dashboard and requests from a separate website for the underlying data used by news outlets and others who design their own maps and graphics.
Government agencies, public health departments, the public and news outlets, including CNN, regularly rely on it for the latest updates on the confirmed cases, deaths and recoveries connected to this harrowing disease.
"We were collecting data on a new virus that nobody understood at a time [when] there was not a single web page dedicated to Covid-19 case count," said Lauren Gardner, the project's chief and an associate professor in, and codirector of, the Center for Systems Science and Engineering at Johns Hopkins Whiting School of Engineering.
From December 2019 to January 2020, Ensheng Dong, Gardner's first-year PhD student, heard from family in China who regularly reported how the outbreak was worsening and upending their lives. After passing an exam that cleared him to earn his PhD, tracking confirmed cases in China was what he wanted to tackle next.
On January 21, over coffee during a weekly research meeting, Dong proposed this idea to Gardner.
His background in spatial data visualization and Gardner's past in modeling infectious diseases converged to create the first iteration of the dashboard — which they finished that night and published the next day. The map's alarming red dots then reflected only 320 reported cases — mostly in China, the rest in Thailand, Japan and South Korea.
"He's gotten to play some pretty major roles in some really central problems to the society for someone that's six months into a degree program," Gardner said. "I keep telling him that he needs to not get used to this and it's not normal. And [that] the rest of his PhD experience is going to be really boring, but I don't think he believes me."
For two weeks from dawn till dusk, Dong lived and breathed the dashboard; it took precedence over his free time outside of class and soon the classes themselves. With Gardner's blessing, he pushed back a required course until the fall semester.
Has maintaining those responsibilities affected his health in any way? "Not really," he said. "I think this is what the research should do. The responsibility for this dashboard gives me energy."
But as cases became globally widespread, they needed help.
"Initially, [I worried about my family]; right now, it's their time to worry about my situation in the US," Dong said.
What was once a modest goal to fill a research gap in a field with antiquated methods for disease tracking is now a tool widely used around the world.
They had no idea "that it would evolve into something that literally impacts almost everybody's life on the planet," Gardner said.
From erratic data to the map on our screens
About 25 people from multiple disciplines now support the dashboard, including graduate students and senior software developers and research scientists primarily based in Maryland, California and England.
Like a lot of people, they're working from home. Their days begin with Zoom calls during which they hash out to-do lists and pressing issues. The conversation continues through a myriad of Slack channels, emails and phone calls.
"It's efficient, but it's boring," Gardner said. "I really miss a bit of the real, in-person world."
Gardner is in charge, so everyone involved reports to her. Her daily to-do list involves running a research group of PhD students, overseeing data additions and design for the dashboard and making strategic decisions for the development of the dashboard. She spends about half of her time on research based on data from the map.
From dozens of sources including local health departments and data aggregating websites, the dashboard reports cases from more than 3,500 locations — at the province level in China; at the county level in the US; and at subnational and national levels elsewhere.
Because they update the dashboard at least hourly, they've had to shift from manual data collection to leaning more on automatic culling — the team praised data wizards at the university's Applied Physics Laboratory for creating a code that periodically travels to trusted websites and scrapes for data. For independent research and the US government, the APL provides technological systems engineering, development and analysis.
The automated code aggregates the data and publishes it into GitHub, a software development platform. An anomaly detection system reviews every number that comes in and holds back anything that doesn't make sense, explained Ryan Lau, a graduate software engineer in the lab. Sometimes the web sources are technologically unreliable, but the case data is accurate in regard to what has been reported, Gardner said.
The teams manually validate and approve numbers before letting them be processed through a geographic information systems tool into the visualizations we see.