Heinrich, PeterHollerer, DagmarKarthaus, ChristophGellrich, MarioRöpke, RenéSchroeder, Ulrik2023-08-302023-08-302023978-3-88579-732-6https://dl.gi.de/handle/20.500.12116/42190Learning to fly an aeroplane is a challenging endeavour; mastering the radio, i.e. talking to air traffic control (ATC) is considered by many students as an especially daunting activity. Using the correct voice procedures and phraseology is paramount for a smooth and safe aircraft operation in busy airspaces. However, training new general aviation pilots in radiotelephony is still a largely manual activity, where students and teachers engage in role-play to mimic the different ends of communication. The design of an aeronautical chatbot, capable of simulating ATC would provide students with plenty of additional training opportunities that they can conduct on their own and at their own pace. A detailed analysis of the voice input would provide feedback on the learning process. This paper reports on the preliminary proof-of-feasibility prototype design as well as synthetic language model training data generation.enAviation EnglishComputer Based Language LearningAutomatic Speech RecognitionText/Conference Paper10.18420/delfi2023-301617-5468