Scott Nicholas
The Defense Advanced Research Projects Agency has awarded an estimated four-year, $8 million contract to Charles River Analytics to develop a software system that can understand, learn and adapt to changes in its technological environment.
CRA plans to develop the Probabilistic Representation of Intent Commitments to Ensure Software Survival system as part of DARPA’s Building Resource Adaptive Software Systems program, the company said Monday.
“We are trying to adapt software used to operate autonomous vehicles,” said Avi Pfeffer, principal scientist at CRA and principal investigator for the PRINCESS project.
“These vehicles need to operate autonomously in dynamic environments, which requires the ability to quickly adapt to changes in the environment.”
CRA will work to integrate machine learning and probabilistic modeling technologies and utilize its research and development activities on unmanned underwater vehicle platforms to build PRINCESS and incorporate capabilities that facilitate response to new operating environments and missions.
The company said it will also provide systems integration, software engineering services, program analysis, verification and other services and assemble a team from Harvard University, the University of Birmingham and the University of Southern California in support of the project.
BRASS aims to realize advances in the design and implementation of software systems that are robust to changes in the physical and logical resources provided by its ecosystem.
