Home > Research > Publications & Outputs > Optimising the trade-off between scheduling and...
View graph of relations

Optimising the trade-off between scheduling and operational delays at congested airports

Research output: Contribution to conference - Without ISBN/ISSN Abstract

Publication date26/04/2019
<mark>Original language</mark>English
Event2nd Joint IMA-ORS Conference - Aston University, Birmingham, United Kingdom
Duration: 25/04/201926/04/2019


Conference2nd Joint IMA-ORS Conference
Country/TerritoryUnited Kingdom


Slot coordination is used at many of the world’s busiest airports as a means of managing scarce runway capacity. The Worldwide Slot Guidelines (WSG) explain how this should be done in order to satisfy the needs of airlines and slot coordinators in a fair and transparent manner. The constraints involved in this process include “declared capacities” of airports, which are often determined at an administrative level based on previous experience.

Mathematical optimisation methods offer the potential to improve upon existing practices for slot allocation, but there are multiple objectives to consider. In particular, a trade-off exists between “schedule displacement”, which quantifies the deviation of a set of allocated slots from the requests originally made by airlines, and “operational delay”, which represents flight delays incurred due to airport congestion. Minimising schedule displacement can often be achieved using integer programming methods, but operational delays depend on stochastic queueing dynamics and must therefore be modelled in a different way.

Many slot allocation problems formulated in the academic literature fail to take sufficient account of the operational delays which are likely to occur as a consequence of high demand rates during peak periods. In this presentation we discuss possible methods for producing slot allocations which achieve an acceptable balance between the conflicting objectives of schedule displacement and operational delay. We discuss the modelling challenges involved and present some of our optimisation strategies, which include iterative methods based on feedback loops between separate tasks and also more direct methods based on the incorporation of queueing constraints in integer programming models.