SciML's Scientific Machine Learning Essential to New Major FDA Project in Artificial Blood

University of Maryland has recently received $46 million dollars in Federal funding through DARPA for the development of an artificial blood product:

a whole blood product, storable at room temperature, that can be used to transfuse wounded soldiers in the field within 30 minutes of injury, potentially saving thousands of lives.

This project involves many groups, including Case Western Reserve University, Charles River Laboratories, Latham Biopharm Group, Ohio State University, and notably Pumas-AI Inc. Pumas has been using using the Julia SciML stack and the techniques developed by the team, such as universal differential equations in order to develop DeepPumas, the first scientific machine learning software in the pharmacometrics industry. Joga Gobburu, CEO of Pumas Inc., mentions in the press release on the funding that:

“We are well-positioned to support this highly complex project that requires the use of advanced modeling and simulation, and a scientific machine learning software system to optimize the prototypes and to test for safety and efficacy in models of complex trauma with multiple complications,”

This press release has made headlines around the world, including with the New York Post, and serves as a proof point for the impact SciML is having on the computational infrustructure of industrial applications.