Abstract
In the mid-term capacity planning process for air cargo, a cargo carrier reserves capacity up to six months in advance for its clients, who provide regular and frequent shipments over multiple flights. We model the underlying capacity allocation problem as a portfolio optimization problem to allocate cargo space on flights while minimizing the demand covariance between allotments and spot market demand. Due to the complexity of the problem, we develop an efficient partitioning algorithm to decompose the problem into subnetworks and cluster demand. The resulting allocation policy is tested using a real world dataset provided by a solution vendor, and it is benchmarked against a risk-neutral allocation policy used in practice. We observed on average revenue improvement by2%, which approximately accounts for$150,000 per week for major cargo carriers.
Original language | English (US) |
---|---|
Number of pages | 25 |
State | Published - 2013 |