APTIMA INC — Department of Defense SBIR Phase II: N07-T031

Description

Despite advances in machine learning for autonomously flying unmanned aerial vehicles (UAV), the intelligence that drives tactical flight path planning is firmly lodged in the heads of human operators. So, too, is the ability to diagnose and resolve faults from automated alerting systems. Recently, an MQ-4C Triton UAV crashed on landing after reporting a rapid cascade of alerts to its Air Vehicle Operator (AVO), who was unable to diagnose the root cause of the Triton’s failure in the wave of redundant and conflicting messages. To address this gap, Aptima and its partner ASEC propose the Operational Partner To Interactively Manage Uninhabited System Procedures & Responses In Multi-Emergencies (OPTIMUS PRIME). As a collaborative, adaptable, cognitive assistant, OPTIMUS PRIME will provide cognitive relief by interactively supporting its human counterpart in understanding and resolving emergency procedures (EPs). Leveraging knowledge about the underlying structure of EPs and air systems, based on prior learning from expert AVOs, OPTIMUS PRIME will improve the effectiveness and efficiency with which AVOs address emergencies during operations. Ultimately, OPTIMUS PRIME will contribute to a decline in the risk of platform loss due to cognitive saturation, confusion, and complexities under high time pressure.