An acoustic analysis of ATC communication

Abstract

Abstract : This report consists of an acoustic analysis of air traffic control (ATC) communications. Air traffic control specialists (ATCS) from a TRACON facility participated in the simulation study. Each ATCS worked light and heavy traffic density scenarios for 2 feeders and 1 final sector. All communications were audio recorded and transcribed verbatim by a retired ATCS. Workload was determined by the number of aircraft under positive control when the ATCS initiated a transmission. Utterances were selected to achieve maximal workload contrast. For each participant, the 5 lowest workload utterances from the Light version of the scenario (simulating that participant's normal work station) and the 5 highest workload utterances from the Heavy version of the scenario (simulating a work station unfamiliar to the participant) were identified and digitized. For all participants, speaking rate (syllables/second), pause frequency (number of pauses/number of words), and pause duration (duration of pauses/number of words) were generated from the selected utterances using the BLISS speech analysis system (Lieberman and Blumstein, 1988). The results indicate that ATCSs tended to pause more frequently and for greater duration under a light workload condition. The hesitations found in their speech may reflect a shift between a more cognitive thinking response mode in light traffic situations where ATCSs know that they have more time to respond and a more automatic mode, which allows them to respond to the increased pace induced by higher traffic loads. In conclusion, it appears that hesitation in speech may be a potential indicator of workload. Despite its highly speaker-dependent nature, hesitation pauses may be a useful indicator of an ATCS's responding in a cognitive, rather than in an automatic mode.