As part of a close collaboration between DLR and the Swiss air navigation service provider Skyguide, our team has analyzed multiple samples of datasets with operational speech recordings from different controller working positions of Skyguide. Just before the ‘Airspace World’ in Lisbon, we can announce that teams of DLR and Skyguide have successfully collaborated to allow the speech recognition and understanding demonstrator to run in shadow mode fed by live air traffic of Zurich.
Our speech recognition and understanding prototype set up includes feeding voice transmissions, push-to-talk signal, and active callsign information from Skyguide systems to the demonstrator of DLR. The demonstrator transcribes the voice transmissions in real time and provides the detected commands, while ensuring that the operational system of Skyguide is not affected in any way. Trained and dedicated staff is subsequently able to rate the system performance for evaluation.
Find the DLR-LinkedIn-Post here: