Intelligent Pairing Assistant (IPA): A Learning-based Framework for Assisting Capability-Need Pairing

ABSTRACT: Within an Air Operations Center, planners working in the Master Air Attack Plan cell make crucial decisions to create the air plan for any given day. They are expected to complete the plan in part by pairing targeting or collection tasks with the available platforms and weapons. Any assistance these planners can get to help create the plan in a timely manner would make the entire process more efficient and effective. One major challenge is orchestrating the target list; designing packages to strike more than one target at a time makes efficient use of resources and improves survivability. Another challenge is making the best use of available resources. For instance, rather than assigning a un-tasked resource to a collection, it may be more expedient to instead assign an aircraft already slated to perform a strike near the target. This proposal outlines a system, the Intelligent Pairing Assistant (IPA), designed to help planners address both of these challenges. The Phase II proposal builds directly on the results of the Phase I investigation, focusing on providing a small number of easy-to-understand pairing recommendations at specific decision points to help planners perform their work more efficiently and effectively. BENEFIT: There are three commercialization paths for the Intelligent Pairing Assistant (IPA) that we will pursue in Phase II: as a MAAPTK plug-in, as an assistant for other AOC systems such as TBMCS, and as a general pairing assistant for other planning domains. The first two options are the primary commercialization paths, pursuing Phase III funding within the US Air Force to transform the system at the end of Phase II into a fully deployed framework for assisting pairing either with MAAPTK or another AOC system. A third commercialization path would be to take the underlying technology and repurpose it for use in Navy Maritime Operation Centers, Department of Homeland Security, or civilian planning systems, for tasks such as wildfire fighting, by encoding knowledge from, and integrating IPA with, the new domain.