A massive collaboration to show how sensors and data streams can detect biological threats and predict future disease outbreaks has advanced with $5 million in federal funding over the next two years.
The NSF Convergence Accelerator approved Computing the Biome for Phase 2 funding in September 2021. Vanderbilt School of Engineering has a lead role in the project, which involves industry and local government, in addition to academia and the federal government.
The collaboration received a $1 million Phase 1 proof-of-concept award in September 2020.
“The Computing the Biome project, which utilizes Microsoft’s Premonition plat- form, is an extremely ambitious pioneering effort aimed at the creation of a global, real-time system for detecting and predicting biological threats as they evolve in the environment,” said E. Bronson Ingram Professor of Engineering Janos Sztipanovits, who is co- principal investigator.
“We are excited to contribute our model-integrated computing technology to this project with the goal of making global impact,” said Sztipanovits, who also is director of the Institute for Software Integrated Systems and professor of computer science.
Using Harris County, Texas—home to the city of Houston and 4.7 million people—as its test case, Computing the Biome aims to demonstrate an artificial intelligence platform that continuously monitors and predicts biothreats. The first target will be mosquito-borne diseases, identifying areas at high risk of West Nile Virus infections.
Microsoft is the big industry partner, providing sensor nodes, species recognizers, models, and industry leadership. Vanderbilt will contribute to the development of open-source data platforms, application design studios, and AI and machine learning algorithms, plus project management.
Computing the Biome has its roots in a project funded by IARPA, the Intelligence Advanced Research Projects Activity, on which Vanderbilt engineers collaborated with Microsoft. The tech giant expanded the project, known as Premonition, led by Ethan Jackson, PhD ’07, a former student of Sztipanovits.
“I am incredibly excited about the integration of academia, public health organizations, startups and established technology providers to tackle important societal challenges,” said Jackson, a senior director at Microsoft. “Delivering solutions requires much more than the fusion of data sets. It requires the fusion of people, ideas and organizations—and this team is going to deliver.”
Scientists need rich and timely data about the distribution and evolution of species in the environment to predict human disease outbreaks. Computing the Biome will build new data streams that combine information such as hyper-local weather, autonomously identified disease transmitting insects, and genetically identified viruses and microbes. Next, the team will develop the AI systems that use those new data streams to detect and predict existing and emerging biothreats.
Tomorrow.io, the second industry partner, provides hyperlocal climate models. The team includes experts from three other universities: Johns Hopkins, the University of Pittsburgh, and the University of Washington.