ADULT SBIR Phase II

The U.S. Army faces challenges to operational test (OT) as a growing number of increasingly complex systems become integral to Multi-Domain Operations (MDO). Collecting data from systems under test (SUT) in field conditions that are similar to real-world operational environments has been a staple of OT for many years. However, the amount and types of telemetry and system data from a SUT, as well as the surrounding Live Virtual Constructive (LVC) wraparound, continues to increase. The Army lacks a means to adequately collect and process this data in an expeditious and meaningful way to feed artificial intelligence (AI) models for system test and test data analysis. Current methods require human-in-the-loop data labeling prior to the data being made available to AI/ML applications. The process is expensive, laborious, potentially inconsistent, and time-consuming. The automated system proposed in this paper offers to expedite this process by reducing the personnel required for labeling data. It also provides the opportunity to label the data in near real-time for immediate use in other systems to achieve additional test objectives, system evaluations, and test data analysis.