Schedule delays are commonly seen in airline operations, given the many operational tasks involved, the stochastic nature of operation time, and unexpected disruptions in operations. In addition, airlines operate in an environment in which airlines have limited control on the constrained system capacities including airport capacities and airspace capacities. The complex interaction between the planned airline schedules and stochastic disruptions may cause flight delays. Delays to a single flight may cause chain reactions in the network in which more flights are affected due to resources (aircraft and crews) and passenger connections among flights. This delay propagation phenomenon is clearer and sometimes more serious for airlines running hub-and-spoke networks than point-to-point networks that provide limited or no passenger interlining services.
The ultimate goal of airline network design and schedule planning is to maintain profitability within limited resources and destinations served in a network. Given the limited resources an airline has, an effective strategy is to adopt the hub-and-spoke network strategy. Hence, an airline establishes a hub (or multiple hubs) to serve a number of spoke cities so to enjoy the economic benefits, namely the economies of density of the network. Airlines adopt the hubbing strategy to various degrees and the level of delay propagation in airline networks often reflects to some degrees the strength of hubbing activities in an airline’s network. Observations of day-to-day airline operations show that the stronger a hubbing network is, the more serious delays may propagate in the network. Although the hub-and-spoke network brings economic benefits to airlines so revenues can be maximised under limited resources, the implications of such a network structure may include: the complexity of resource connections, passenger connections among flights, longer travel time due to indirect flights, and the vulnerability of airline schedules under stochastic disruptions, especially at hubs.
A route-based analysis is designed to investigate the reliability of planned aircraft routing schedules. A ‘route’ is a sequence of flights operated by the same aircraft during a routing cycle. Each arrow represents a flight from an origin airport to a destination airport. Ground time is scheduled in between flights at airports for aircraft turnaround operations including processing check-in and connecting passengers, baggage, cargo and refuelling for a following flight by the same aircraft. Airline data are used to calculate the total daily inbound delays (DDm), the total daily outbound delays (DDm), and the total daily operational delays (DOpD) of all flights in the network. The network-wide inbound delays reveal how delays propagate among flights via aircraft routing in the network and whether daily and seasonal effects are significant on airline schedule operations. The network-wide outbound delays show how ground buffer time has helped reduce the delay propagation among flights with the existence of stochastic operational delays from aircraft ground operations. Comparisons between the domestic routes operated by the anonymous carrier in 2004 are conducted and results are shown in Figure 1below.
It can been seen in Figure 1 that the delay propagation in the network follows a pattern which shows that the total daily inbound delays (DDm, the solid line) of the network are close to but higher than the total daily outbound delays (DDn, the dotted line) due to the designed ground and enroute buffer time in the schedule. We can also see that the total daily operational delays (DOpD, the broken line) of the network are high. These operational delays have been partially compensated by buffer time designed in the schedule, so the gap between DDm and DDn line remains narrow as we see in Figure 1.
The implications of results from this study are twofold. First, the results advance our understanding of the complexity embedded in airline network design and its implicit influence on airline operations. The complexity comes from the goal of maximising passenger intake in the network and a possible consequence of this complexity is to compromise the reliability and robustness of airline operations. Secondly, this understanding can improve the effectiveness of schedule optimisation methodologies, which aim at improving the robustness of airline schedule planning and operations. Based on those fundamental issues addressed in this paper for designing airline timetables, the subsequent schedule optimisation task following schedule planning could achieve a higher robustness level by strengthening the feedback loop between airline schedule generation, aircraft routing optimisation and schedule operations.
This project is supported by the Australian Research Council (ARC) Discovery project (2005-2007), Faculty of Science, UNSW and an anonymous airline.
Any comments to this work are highly welcome. Please contact Dr. Richard Wu at firstname.lastname@example.org.
This paper is under peer review by the Journal of Air Transport Management. Copyright is reserved.
More articles containing information about research in the Department of Aviation can be found on the homepage articles web-page.