DARPA Chooses Research Teams to Develop Models and Simulators for Data Management

The Defense Advanced Research Projects Agency has selected nine research teams that will participate in the Automating Scientific Knowledge Extraction and Modeling program.

The teams will use artificial intelligence approaches and tools to develop complex models and simulators for data and knowledge management. They will apply the models and simulators to various use cases with a focus on epidemics and space weather.

ASKEM is split into two phases, with each phase lasting 21 months. The first phase will focus on accuracy, timeliness and maintainability while the second phase will focus on scalability, DARPA said.

The program will span four major technical areas, namely the development of technology to extract and link knowledge from diverse inputs; the development of approaches for composing, decomposing, comparing and sustaining models; the development of tools for creating complex model simulators; and the creation of a visual meta-modeling platform.

Teams from the University of Wisconsin, the University of Arizona, Lum .ai, SIFT and the University of Utah and the Massachusetts Institute of Technology will work on the first focus area while researchers from the University of Florida, the University of Colorado Denver, Topos Institute and Harvard Medical School will focus on the second area.

MIT, the University of Pittsburgh, Pacific Northwest National Laboratory, Basis and the University of Texas at Austin have been assigned to the third focus area while Uncharted Software and Jataware will work on the fourth focus area.