A network science approach to identify disruptive elements of an airline Vinod Kumar Chauhan 1,2* , Anna Ledwoch 1,3 , Alexandra Brintrup 1 , Manuel Herrera 1 , Vaggelis Giannikas 3 , Goran Stojkovic 4 , Duncan Mcfarlane 1 1 Institute for Manufacturing, University of Cambridge 2 Department of Engineering Science, University of Oxford 3 School of Management, University of Bath 4 The Boeing Company November 16, 2022 Abstract Nowadays, flight delays are quite notorious and propagate from an originating flight to connecting flights, which lead to big disruptions in the overall schedule. These disruptions cause huge economic losses, affect the reputation of airlines, lead to a wastage of time and money of passengers, and have a direct environmental impact. This paper presents a novel network science approach for modelling and analysis of an airline’s flight schedules and its historical operational data. The final aim is to find out the most disruptive airports, flights, flight-connections and connection-type in an airline network. In this regard, disruptive elements refer to influential or critical entities of an airline network. These are the elements which either can cause (as per airline schedules) or has caused (as per the historical data) the biggest disturbances in the network. An airline, then can improve their operations by avoiding disruptive elements. This can be achieved through introduction of an extra slack time between connecting flights and by creation of alternate arrangements for aircraft and crew members for the disruptive flights and flight-connections. The proposed network science approach for disruptive elements’ analysis is validated with a case-study of an operating airline. Interestingly, the analysis shows that (potential) disruptive elements in the schedule of the airline are also (actual) disruptive elements in the historical data and should be attended first to improve operations. Keywords: air transport, flight delays, airline disruptions, delay propagation, network science. 1 Introduction Nowadays, flight delays occur quite frequently and it has been observed that in 2017, 20% of flights arrived late by at least 15 minutes or more in Europe (Walker (2017)). These delays can be classified, mainly into three categories: airline issues, airport issues and weather issues. The airline issues arise due to problems at the airline’s end, e.g., delay in passenger boarding and disembarkation, aircraft repairs and sudden un- availability of crew members etc. The airport issues arise due to problems at the airport authority’s end, e.g., unavailability of slots for flights to take off or land, longer time in security checks and airport closures etc. The third category is the severe weather conditions, which can’t be controlled like the other two, e.g., storms and snow falls can disturb the airport operations (Brueckner et al. (2022)). Airline operations represent a complex distributed transportation system, which have several interacting and inter-dependent entities, like passengers, crews, airlines and airports etc. So, when one thing goes wrong, it has the potential to affect the whole system and because of that when one flight is delayed, * vinod.kumar@eng.ox.ac.uk (This work was done at University of Cambridge) 1 arXiv:2211.08140v1 [physics.soc-ph] 19 Oct 2022