Synchronised Scheduling of ULD Breakdown and Shipment Pickup with Slot Confirmation in an Air Cargo Terminal
M.J.W.G. van Hugten (TU Delft - Aerospace Engineering)
A. Bombelli – Mentor (TU Delft - Operations & Environment)
T.R.J. Helsdingen – Mentor (KLM Royal Dutch Airlines)
I. de Pater – Graduation committee member (TU Delft - Astrodynamics & Space Missions)
D. Zappalá – Graduation committee member (TU Delft - Wind Energy)
A.B.A. Lubbe – Mentor (KLM Royal Dutch Airlines)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Air cargo is a vital component in the worldwide supply chain and economy. In the KLM Ground Handling (GH) terminal, Unit Load Device (ULD) breakdown and shipment pickup are scheduled separately, with limited consideration of storage capacity and truck delays, which results in inefficiencies in schedules and higher storage requirements. To counteract these issues, this study proposes a scheduling model which synchronises ULD breakdown and shipment pickup and determines the effect of a truck slot confirmation system where drivers must confirm their presence some time before their scheduled slot, called the confirmation horizon. A Mixed-Integer Linear Programming (MILP) model extended from cross-docking literature is considered. Moreover, a Genetic Algorithm (GA) is developed to improve computational complexity. It is extended to a rolling horizon (RH) implementation, called the RH-GA, which can react to discovered delays. Using the baseline MILP (b-MILP) a comparison to the current scheduling method is performed. Moreover, experiments on realistic KLM GH scenarios are carried out using the RH-GA. The MILP failed to find feasible solutions within 8 hours for many small/mid-size scenarios. The GA always reached feasibility in tested scenarios and its runtime scaled linearly with scenario size. In the few scenarios which could be benchmarked, the synchronised GA outperformed the b-MILP. Moreover, introducing a non-zero confirmation horizon significantly improved the objective compared to no confirmation horizon. Across tested cases, mean peak inventory decreased by 9% and total tardiness by 2-3% relative to the 0-hour horizon case. Synchronised scheduling with a slot confirmation system can reduce storage requirements and improve on-time performance at the KLM GH terminal, indicating that further development and implementation of both the model and confirmation system may be valuable.
Files
File under embargo until 16-01-2028