Optimizing Pricing and Packing of Variable-Sized Cargo

Organizations have successfully used dynamic pricing to optimize revenues for many years, where research and practice have mainly focused on applications with independent discrete commodities, for example an airline ticket. In this research we consider applications where the commodity cannot be considered independent and the value of the commodity available to sell depends on the combination of previously accepted demand. In this paper we focus on vehicle ferries, where the accepted vehicle bookings are packed on the ferry to leave a useable space for future bookings. Certain combinations of vehicles may leave areas of unusable space therefore affecting future revenue. While this application is the focus of the paper, there are numerous industries that face similar challenges including freight and the sale of advertising on television and radio. In this paper we present a method that simultaneously solves the pricing and resource utilization problem with a discrete set of product types and stochastic demand. Our approach combines a dynamic pricing model, with a mixed integer linear program to optimize the packing. We present results for real-world examples from the ferry industry and discuss extensions to the method to improve the selection of vehicle configurations.