An expert system to manage call sign similarity for safer air traffic operations

The continuous growth of air traffic and radio-telephony procedures led to a large increase of messages exchanged between pilots and air traffic controllers. At the same time, this numerosity brings inherent potential for larger incorrect flight identifications, with consequence on both safety and efficiency in terms of air-ground communications. This paper focuses on the Call Sign Similarity phenomenon, i.e., an event where two flights share a similar Call Sign. Over the last two decades, EUROCONTROL has been trying to find a systemic solution to this problem and, in this regard, has drafted a set of Call Sign Similarity Rules to help airlines to deconflict their schedules before actual operations. This work details an expert system built to deal with selected Call Sign Similarity deconflicting rules and shows a data mining approach for their application to planning data. Exemplary results are proposed to prove the feasibility of the approach and to motivate the importance of Call Signs similarity detection.