Identifizierung leistungsrelevanter Merkmale der Teamkommunikation in der Domäne Flugsicherung

Kurzfassung

Work conducted by teams influences the performance of the air traffic system. This thesis proposes the analysis of team communication to assess teamwork und understand its influence on novel air traffic management concepts in high-fidelity simulations. The influence of the situational factors experience and communication media on communication behaviour was studied. Tasks of en-route air traffic controllers were used as an example to identify performance enhancing behaviours. First, they were defined based on theory and task analysis and integrated into a frame model of team communication. Second, these behaviours were reviewed by using empirical data. Therefore, the simplified air traffic control team task TeamCom was developed where team members had to jointly monitor air traffic and prevent separation losses. To capture communication behaviour, metrics were derived from speech data. Two studies with novice samples of 29 dyads each were conducted which analysed the influence of communication media and experience on team communication. Results show that communication was rarely systematically influenced by experience, but significantly by the time needed for using communication media. Furthermore, in these studies teams with good and poor performance could be discriminated by metrics that related timing of communication to requirements and pace of the team task. Teams with good performance exchanged relevant information for their tasks earlier and communicated situations earlier, which required joint actions from the team. Additionally, high-performance teams communicated their intentions and gave each other more feedback. Based on these findings, added value, limitations, and informative value of communication analysis are discussed. As unexpected performance-related communication behaviours could be identified, team communication analysis can contribute to the refinement of theoretical models of teamwork and team performance in complex human-machine-systems. To enhance practicability of the approach, a perspective for automating the analysis is given.