Researchers Needed A Rapid Scaling Data Management Platform For Responding To COVID-19 Project With U.S. Military

MIT & UC San Francisco (UC TemPredict)

Challenge

During COVID-19, UCSF, UCSD and MIT partnered with the US Navy and US Army to build cutting-edge pandemic prediction models. The vast volume of data collected from IoT devices (wearables) used by first responders highlighted the need for scalable systems to collect and analyze big data in a private and secure way.

Solution

In partnership with Sherlock, the TemPredict project stood up a secure and scalable data infrastructure and created data products to gather continuous physiological data for COVID-19 detection algorithm development. Without these data products the predictive models would have been limited with fewer dimensions and variables, and thus be less accurate.

Result

The project team was able to quickly get started and respond to urgent needs of the participating researchers. The project collected data from roughly 50,000 people and generated personalized and population-based models for continuous risk assessment, and ultimately will help save lives by predicting future outbreaks. Sherlock’s expertise helped in managing the data and algorithms to be maintained to both HIPAA and CUI requirements.

“Sherlock was able to quickly get started and respond to urgent needs of the participating researchers. The project collected data from roughly 50,000 people and generated personalized and population-based models for continuous risk assessment, and ultimately will help save lives by predicting future COVID-19 outbreaks. Sherlock’s expertise helped in managing the data and algorithms to be maintained to both HIPAA and CUI requirements”

_{Ben Smarr
Assistant Professor
Bioengineering and Data Science | UC SanDiego}