Optimizing Air Cargo Load Planning and Analysis

This paper describes our work in using mathematical optimization to assist air cargo load planning for an air cargo company, one of the world's top 10 air cargo carriers. The airline handles roughly 80 to 90 thousand tones of cargo each month. With this volume of cargo, even the slightest improvement in the load planning process will have a major impact to overall performance and efficiency. In this project, we developed a Web-based application to firstly perform long-term forecasting based on an analysis of historical data and then secondly operational load planning with mathematical optimization. Forecasting is important as the data will assist the airline in drafting its long-term contracts and determination of allotments. Load planning, on the other hand, is for operational term planning of how ULDs (Unit Load Devices – air cargo containers/pallets) are loaded into aircrafts. This process is usually done roughly 2 hours before departure when all the details of the cargo are present. Therefore, the performance of the application is crucial. In this project, in order to help users to get more data dimensions to analyze, a data warehouse was built. Load planning was performed using MATLAB and Mosek