TANDAMS (Teamwork ANalysis and Development in Analyst-Machine Systems)

Intelligence analysts increasingly rely on machine learning (ML) capabilities to process an ever-expanding set of sensor data – significantly multiplying the amount of data available to analysts and increasing the risk for information overload. Despite this proliferation of information, little attention has been placed on ML capabilities that support analysts as they draw conclusions, link information, and refine intelligence products. Based on results of on a previous validation study to analyze the impacts of increased use of machine learning (ML) within intelligence processing domains, SoarTech proposes the development of TANDAMS, a capability that will enable intelligence analysts to more effectively integrate observations produced by ML algorithms, particularly when information must be combined across formats (e.g., linking textual HUMINT data to activities recognized in full-motion video feeds) to satisfy collection requirements. In this Phase I, SoarTech will work with AFRL and operational stakeholders at the Air Force’s Distributed Common Ground System (DCGS) to develop proof of concept demonstrations and solicit feedback from capability stakeholders.