Task Examination/Aggregation/Learning (TEAL) to Support Adjustable Autonomy

Future Combat Systems are envisioned as a networked system of systems that will integrate tomorrow’s soldier with families of manned and unmanned platforms and sensors. However, tele-operated and pre-FCS legacy systems will need to be upgraded to be sophisticated and “intelligent” enough to interact within FCS. A practical pathway to achieve such upgrades is to add the required functionality into Operator Control Units (OCUs), leaving the vehicles themselves unchanged while providing operators with high level commands that better match desired tasking. In Phase I, the Soar Technology team designed, implemented, and demonstrated an initial version of the Task Examination, Aggregation and, Learning (TEAL) concept to develop higher-level commands from observed sequences of existing unmanned vehicle commands, incorporating user command validation to ensure that newly added commands are useful. For Phase II, we will transform the TEAL approach demonstrated in Phase I into a robust, independent reasoning software module with a well-defined application programming interface (API) that can be usefully and easily integrated into a wide variety of OCUs, produced by multiple OCU vendors, and will demonstrate the TEAL functionality used in the control of physical robots.