Milind G. Sohoni

Robust Airline Scheduling Under Block-Time Uncertainty

Airline schedule development continues to remain one of the most challenging planning activities for any airline. An airline schedule comprises a list of flights and specifies the origin, destination, scheduled departure, and arrival time of each flight in the airlines network. A critical component of the schedule development activity is the choice of flight block-times, which depend on several factors. Many airlines decide schedule block-times based on fixed percentiles of block-time distributions built from historical data, however, such techniques have not resulted in significantly improved on-time performance (OTP) of the schedule during operations. Thus, from a passengers perspective, the service-level guarantee of an airlines network continues to be low. We first define two service-level metrics for an airline schedule. The first one is similar to the OTP measure of the U.S. Department of Transportation and we define it as the flight service level. The second metric, called the network service level, is geared toward completion of passenger itineraries. We then develop a stochastic integer programming formulation that optimally perturbs a given schedule to maximize expected profit, while ensuring the two service levels. We also develop a variant of this model that maximizes service levels, while achieving desired network profitability. To solve these models, we develop an efficient algorithm that guarantees optimality. Through extensive computational experiments, using real-world data, we demonstrate that our models and algorithms are efficient and achieve the desired trade-off between service level and profitability.

Operational airline reserve crew planning

Airlines are continually faced with the challenge of efficient utilization of their cockpit crew resources. In addition to regular flying crews, some airlines have to maintain significant reserve staffing levels to meet contractual obligations and provide smooth daily operations. Reserve crews are required to cover trips remaining unassigned due to disruptions during daily operations. Airlines using a bidline system to award crew work schedules require additional reserves to cover scheduling conflicts, which result in trips dropping out of optimized bidlines. Whenever reserves are unavailable to cover these trips during daily operations, the airline has to pay a premium to cover these trips using regular pilots. The resulting operating expenses can be significant. Furthermore, inefficient utilization of reserves can cause excessive long-range crew staffing resulting in additional training and new hire expenses. In this paper, we propose a new optimization strategy to increase reserve crew utilization and build monthly reserve crew work schedules by addressing the issue of scheduling conflicts and daily operational reserve requirements.

A dynamic neighborhood based tabu search algorithm for real-world flight instructor scheduling problems

The multi-objective flight instructor scheduling problem is an optimization problem that schedules instructors to teach a set of pilot training events. The objectives of the problem are to minimize labor cost, maximize workload consistency and maximize flight instructor satisfaction of their assignments. The problem is further complicated by various hard and soft constraints. We study a multi-objective cost function and convert it to a scalar-weighted objective function using a priori weighting scheme. We then design an efficient dynamic neighborhood based tabu search meta-heuristic to solve the problem. The algorithm exploits the special properties of different types of neighborhood moves. We also address issues of solution domination, tabu short-term memory, dynamic tabu tenure and aspiration rule. The application of the algorithm in a major US airline carrier is reported and the results show that our algorithm achieves significant benefits in practice.

Delta optimizes continuing-qualification-training schedules for pilots

Delta Air Lines periodically trains its cockpit crew members to maintain their flight qualifications. The negotiation of the new pilot working agreement (PWA) in June 2001 introduced a short and stringent planning cycle and a new training-pay structure that affects operating costs. The downturn in airline business after September 11, 2001 forced Delta to reduce its workforce and to modify its requirements for scheduling pilot training. To mitigate Deltas exposure to costs and to automate the scheduling process under a rigid planning time line, we developed and deployed an automated optimization system, CQOPT, that builds and assigns training schedules based on individual pilots requirements. The primary objectives of CQOPT are to minimize overall operating costs and maximize training assignments. Shortly after CQOPTs deployment in May 2002, the planning cycle dropped from several days to a few hours. Delta expects to save

Payment for Results: Funding Non-Profit Operations

7.5 million in annual operating costs by using CQOPT to schedule continuing qualification (CQ) training for its pilots.

Differential Pricing for Information Sharing Under Competition

The non-profit sector (NPO) faces two critical challenges in delivering public goods effectively and raising funds for developmental projects. The first challenge arises due to donors being uncertain about the NPOs efficiency to deliver the expected output (information asymmetry), and the second because the actual benefit provided is subject to uncertain exogenous shocks (outcome uncertainty). To partially address these challenges, the sector has seen the growth of the payment for results (PfR) funding approach, where the donor funds the NPO after the project is implemented and the outcome is observed, compared to the traditional funding (TF) approach where the donor contributes ex-ante. While the PfR approach addresses the donors challenge, the NPO still faces a significant risk of shortfall in funds. In this paper we model the interaction between the donor and a NPO as a Stackelberg game and characterize the conditions under which TF and PfR emerge as the equilibrium outcomes when there is information asymmetry and outcome uncertainty. The donor moves first and decides the donation amount if the NPO were to opt for TF and the NPO responds by choosing TF or PfR. Counter to initial intuition, we find that the donor does not prefer PfR under all conditions. We find that incentivizing the NPO to choose TF increases the donors expected utility when the negative impact of uncertainty and the donors willingness to contribute are high. On the other hand, the donor prefers the NPO to use PfR when the impact of outcome uncertainty is relatively small. Using numerical studies, we find that PfR does not always maximize the expected benefit provided. We find that the gap between the maximum possible expected benefit and expected benefit at equilibrium is significant when outcome uncertainty and donors willingness to contribute are high.