Automatic air cargo selection and weight balancing: A mixed integer programming approach

•The paper describes cargo selection and weight balancing for freight carriers.•It presents a mixed integer programming model enabling decision support.•A wide range of stability, safety and operational constraints are considered.•The model is validated on a representative set of flights of a commercial carrier.•Testing shows an improvement on both selected cargo value and aircraft balancing.

The present contribution introduces a mixed integer linear programming model as a decision support tool for air cargo load planning. The main objective for the model is to find the most profitable selection from a set of cargo to be loaded on an aircraft. The secondary objective is to minimise the deviation between the aircraft’s centre of gravity, and a known target value so as to reduce fuel consumption and improve stability. The model is subject to a large number of constraints that ensure structural integrity and stability of the aircraft, as well as the safety of the cargo and crew. A set of additional constraints guarantees safe and efficient loading and unloading. Experimental results on real-life data show that the model outperforms human expert planners on both objectives, while remaining computationally fast enough for interactive use. This advocates the use of such a decision support model for all air cargo load planning.