Summary
Google says two Earth AI research prototypes helped public-health teams map communities at risk during an Ebola outbreak in the Democratic Republic of Congo. The company also describes research using population and environmental data to forecast several health trends.
Google says two Google Earth AI research prototypes helped public-health teams map communities at risk during an Ebola outbreak in the Democratic Republic of Congo (DRC) and model how the disease might spread. The company also describes research applying its population and geospatial models to cardiovascular mortality, immunisation, dengue and cholera. The work combines local health information with environmental signals, mobility data and satellite imagery to help address gaps in conventional reporting.
Prototypes used during the DRC Ebola outbreak
The World Health Organization’s Regional Office for Africa (WHO AFRO) Emergency Preparedness and Response Hub and the DRC’s National Institute of Biomedical Research (INRB) used two research prototypes: a Geospatial Reasoning agent for conversational mapping, and a planetary prediction engine for autonomous disease forecasting.
Google says the WHO AFRO team used the mapping agent to identify remote mining corridors with high exposure risk and human mobility. In minutes, it mapped 48 exposed settlements and more than 45,500 people considered at risk—a process the company says would normally take weeks. The mapping helped local responders plan mobile laboratory deployment and coordinate border surveillance.
Working with INRB’s Epidemic Modeling and Intelligence Unit, Google also built models to estimate the risk of Ebola spreading into areas without reported infections. The weekly forecasts combine mobility flows, historical case trends and Earth AI models and datasets, giving coordinators planning time before cases arrive, according to the announcement.
These tools address different parts of outbreak response: mapping can help locate potentially exposed communities, while forecasting can help officials anticipate where further spread may occur. Both are research prototypes, rather than a single public-health service described as generally deployed.
Population data for longer-term health planning
Google describes its Population Dynamics Foundation Model (PDFM) as combining aggregated search trends with mobility and environmental patterns to create a high-resolution picture of communities. Partners can integrate its insights into their own systems. The company says this type of population information can complement existing health data and support predictions where routine reporting is incomplete or delayed.
In work with NYU Langone Health, researchers used real-time population signals in models projecting same-year cardiovascular mortality. Google reports performance similar to conventional approaches, and in some cases better. Researchers at Mount Sinai Health System and Boston Children’s Hospital used cross-border behavioural patterns to improve estimates of vaccination rates in US counties for measles, mumps and rubella.
For dengue, researchers at the University of Oxford and Tecnológico de Monterrey combined PDFM with local climate models to forecast outbreaks across Mexico, which Google says improved forecast accuracy. In DRC cholera evaluations with WHO AFRO, combining epidemiological records with PDFM improved identification of outbreak-prone health zones up to eight weeks in advance.
The announcement describes these results qualitatively and does not provide sample sizes or numerical performance measures for the evaluations. That makes the reported improvements difficult to compare across diseases or with specific conventional models based on this account alone.
PDFM embeddings are commercially available in Preview as Population Dynamics Insights, a geospatial dataset from Google Maps Platform. Researchers can request no-cost access for select use cases, and eligible organisations can apply for Google Earth credits through Google Maps Platform Public Programs. Google.org has also provided funding to INRB for modernising local testing and disease surveillance.