Advanced Learning Through Human-Enhanced Automation

The subject of this Phase II proposal is ALTHEA (Advanced Learning Through Human-Enhanced Automation), an innovative system for human-machine knowledge integration that that allows analysts to provide expert feedback to automated data processing systems, thereby improving mission performance. As space emerges as a contested domain, the reliable and timely generation of indications and warning (I&W) is becoming increasingly critical. Many elements of the space defense mission rely on automated data processing algorithms in order to produce timely I&Ws. While data and information utility are critical, the data itself is often noisy or incomplete. As a result, machine learning algorithms may fail or produce results with insufficient confidence levels. To mitigate these issues, ALTHEA integrates human- and machine-generated knowledge by engaging expert analysts as a resource by identifying high-impact data uncertainties for them to resolve. By soliciting human expertise in this way, ALTHEA improves data utility and downstream machine-learning algorithm performance, thereby improving the quality of I&Ws for space situational awareness (SSA) and space battle management command and control (BMC2) applications when incorporated into the appropriate analyst workflows.