Using publicly available health data, the software known as "PRIME" is able to make predictions on a county-by-county level. Amid the most recent pandemic, the predictions helped the New Mexico Department of Health in its response to the spread of COVID-19, according to Sandia officials.
Sandia is headquartered in New Mexico and has a campus in Livermore next to the Lawrence Livermore National Laboratory.
With the pandemic tracking software's potential to prevent the spread of an outbreak, the Department of Energy has invested funding to support further developing the technology.
The Sandia researchers have teamed up with Illinois-based Argonne National Laboratory to advance their work toward learning to stop outbreaks before they have a chance to spread. Their joint proposal was accepted by the DOE's Office of Science last August, with a total project award of $825,000 a year for two years.
Jaideep Ray, one of the researchers behind the software, told the Weekly in a recent interview that developing the tool has actually been a project 18 years in the making, dating back to the anthrax scare in the early 2000s when a powdery substance containing the infectious bacteria was being distributed through envelopes in the mail.
"That's when this whole business started of looking at any incomplete data that we get today and trying to predict how bad things are going to be -- predict how many people are still hidden or how quickly the disease is spreading," Ray said.
In the years since then, the team -- which includes fellow researcher Cosmin Safta -- has continued refining the technology, researching and collecting data. However, it was in 2020 when the COVID-19 pandemic hit that all of the theory and mathematics they had developed was revamped into a software in a span of two weeks, which allowed the researchers to begin conducting forecasts for California and other states throughout the country.
With this new DOE funding, Ray said there are additional refinements and improvements to be made to the existing software.
"The thing that we really want to do is not just forecast what the disease is going to do but also allow a public health official to plan out various kinds of scenarios. For example, if I imposed certain kinds of lockdowns today, how would the disease then evolve and what kind of economic and social consequences is it going to have?" Ray said, explaining how the software could offer even more insight ahead of a future outbreak.
"You can do this using models and there are models to do that but you want those models to be at least predictive and getting those models to be predictive -- to agree with data -- is an extremely tough job," he added.
Where Argonne and Sandia come together in achieving this goal is that Argonne has the agent-based modeling tool Ray described and Sandia has the mathematical framework. "The DOE said, the two of you work together and get this done and so, we're getting it done," Ray said.
"Such a breakthrough, funded through this research grant, might one day help the nation avoid the debilitating and deadly impacts of the next pandemic," Sandia officials said in a news release.
