Yuri Levin, Tatsiana Levina, Jeff McGill, Mikhail Nediak and Huseyin Topaloglu

COIN-OR Applications in Air Cargo Capacity Management and Dynamic Pricing

Revenue management and related dynamic pricing practices are gaining popularity in different industries and have engendered a growing body of academic research in recent years. However, the general area of capacity and revenue management for the supply chains has been developing significantly slower. This slower pace is due, in part, to the computational complexity of cargo transportation capacity and RM problems as well as an inherent complexity of decisions under imperfect information. Another difficulty is a transfer of academic results, which often employ complex computational algorithms, to industry.

The use of open Operations Research software technologies such as COIN-OR improved the portability of algorithms and helped in their practical application. The team successfully applied COIN-OR technologies in to their research in both online learning and cargo capacity management.

Online learning is an important tool to help companies deal with the uncertain and dynamic nature of modern business environments. The methodological foundation of their research is the general aggregating algorithm approach to online learning. Article [1] employs IPOPT and DFO technologies in online learning of strategic consumer behaviour in the context of dynamic pricing.

Their research in air cargo capacity management used COIN-OR technologies CBC and CLP. In [2], they proposed a stochastic dynamic booking/shipping control model for air cargo capacity management on a network. The model captures air cargo-specific uncertainties in the booking arrival process, capacity of flights, and capacity requirements of shipments. The model explicitly captures the periodic nature of the flight schedule which distinguishes it from existing stochastic dynamic models in network passenger RM. In [3], they presented a model that integrates multiple allotment contracts and spot market activities of an airline for a group of parallel flights. They constructed high quality accept/reject policies for the booking requests that occur on the spot market, and showed how to obtain an upper bound on the optimal total expected profit from the spot market. This upper bound provides a spot market valuation tool that facilitates allotment contract negotiations and allows the airline to conduct a combinatorial auction of its cargo capacity.

It is with great pleasure that we recognize this work with the COIN-OR INFORMS 2009 Cup.

[1] Levina, T., Y. Levin, J. McGill and M. Nediak (2009). Dynamic Pricing with Online Learning: Application to Markets with Strategic Consumers, Operations Research, Vol. 57, No. 2, pp. 327-341.
[2] Levina, T., Y. Levin, J. McGill and M. Nediak (2010). Network Cargo Capacity Management, forthcoming in Operations Research.
[3] Levin, Y., M. Nediak and H. Topaloglu (2010). Cargo Capacity Management with Allotments and Spot Market Demand, under review in Operations Research.