Air Cargo Handling Demand Forecasting; 'As a support tool for short-term decision on manpower deployment at World Port'
This item has file attachments that are restricted and can only be viewed from the TU Delft campus network.
This item has an embargo placed on the file attachments in effect until: 2013-11-12.
KLM Cargo would like to improve short-term resource scheduling at their World Port cargo handling terminal, in order to increase both quality and efficiency of their service. Forecasts of air cargo demand could support decisions on manpower deployment. This thesis contains a forecast comparison study for short-term forecasts of the amount of departing cargo from World Port in kilograms per day, based on historical booking data. A single regression forecast model on pre-departure booking levels came up as top-performing model, capable of predicting cargo demand per day with only 3% mean absolute percentage error. This thesis also examines departure behaviour over the course of each day of the week. Furthermore, arrival behaviour is analysed, which is defined as the connection between the departure flow and several arrival flows at the terminal. The main conclusions out of the behavior analyses is that cargo departure is unevenly spread out over each day and that cargo is delivered at World Port relatively far in advance of planned departure.