Robust Voice Communication Understanding for Single-Pilot Operations

Motivated by the anticipation of pilot shortage in the future and the quest of cost reduction in airline operations, the single-pilot operation (SPO) concept emerges as a promising alternative of the current-day multi-pilot operations. This proposal addresses the feasibility of constructing a spoken language understanding system for decoding voice communications in Air Traffic Control. In particular, we address the issue of developing a voice communication understanding system (VCUS) that would serve as one key component in both cockpit automation and ground-based automation for supporting the SPO concept. Leveraged from our prior development on noise-robust speech recognition system for the Navy and virtual agents for NASA to support human-in-the-loop simulations, an infrastructure of VCUS in Air Traffic Control of commercial flights will be developed. A feasibility demonstration of the VCUS that extracts out semantic information for persistent display of clearance message within the flight deck will be provided by the end of the Phase I research. Phase II work will utilize the infrastructure built in Phase I to expand the VCUS to a full-scale prototype that supports cockpit automation and ground-based automation for SPO.