Enhancing Reduced-Crew Flight Operations with NLP-Driven Real-Time Dynamic Replanning

Real-time dynamic flight plan updates are a common task required of pilots in wide variety of Air Force and Special Operations missions, such as humanitarian relief (airdrop), deploying aerial munitions, and aerial refueling operations. In these missions, the pilot is tasked with verbally coordinating “last mile” type flight updates (with other aircraft, controllers, ground units, or refueling receivers) to avoid weather, update rendezvous waypoint location/timing, etc. Coordinating these flight updates is a burdensome and task saturated effort that must be solved to allow reduced crew operations in DAF / SOCOM fixed wing aircraft. Utilizing Merlin’s commercially developed dynamic replanning tool and our artificial intelligence based Automated Communication system, the Merlin Pilot (MP) can reduce crew workload during dynamic replanning operations for pilots across the DAF and USSOCOM. Merlin’s approach to this problem set follows current TTPs while taking a step towards reduced crew flight operations. Our solution architecture to “last mile” dynamic replanning is platform agnostic, enabling broad use of this technology for military aircraft and use cases. Additionally, our solution supports inputs from different communication modalities, so future iterations of this technology can consume data link inputs to operate in contested/emission-controlled environments. Merlin proposes utilizing the aerial refueling use case as a first concept of operations demonstration, allowing us to leverage existing data gathered from ongoing work with Air Mobility Command and AFMC. Applying this technology to the operationally deployed aerial refueling use case, the receivers verbally request an updated rendezvous waypoint from the MP. The MP then formulates a new flight plan to the updated destination. Lastly, MP communicates the updated flight plan to the controller, who communicates back for confirmation. Upon confirmation, the MP will load the updated flight plan into the Flight Management System. Automated real-time dynamic replanning can reduce crew workload and provides a necessary step towards reduced crew capable aircraft.Merlin Labs proposes to address this critical need for real-time dynamic re-planning with a system consisting of:An optimization algorithm to suggest new updated flight paths for an individual tanker based upon the receiver's requested rendezvous waypoint and any other constraints (e.g. restricted airspaces, adverse weather conditions);Natural Language Processor (NLP) - to monitor airborne communications to recognize receiver to ownship and ownship to controller communications and identify actionable requests for updates to their rendezvous waypoint;Response Generator to generate requests for updated flight plan over radio voice to the controller and generate approvals/denials for updated rendezvous waypoints to the receiver.