Khaled Abdelghany, Southern Methodist University

Vitaly Guzhva, Embry-Riddle Aeronautical University

This research presents a modeling framework to provide a solution to the airlines' scheduled block time (SBT) design problem. The framework implements a novel data-driven optimization model that considers reliability measures at both the flight and the schedule levels. The framework enables airlines’ schedule planners to examine the relationship between the allocated SBT for each flight and the aggregate schedule reliability and make informed decisions regarding the trade-off between schedule reliability and operational costs. The efficacy of the proposed framework is examined using the schedules of two major U.S. airlines. Experiments are conducted to evaluate the model's effectiveness. In these experiments, the SBT estimated by the model is compared against the SBT provided by the airline for each flight in the schedule. In addition, both the estimated SBT and the SBT provided by the airline are compared against the Actual Block Time (ABT) recorded for all flights in the schedule. Results confirm that there is a room of improving airlines' current practice of block time estimation.

Soheil Sibdari, University of Massachusetts Dartmouth

We use international air-service historical data combined with economic, tourism, and trade data sets to examine the impact of non-stop Trans-Atlantic flights between the U.S. and selected African countries on local trade. We use quarterly data between 2009 and 2019, collected from public and private resources and use a panel regression method to conduct and illustrate our results. The outcome of this research helps policy makers, commercial airlines, and international organizations to make better capacity investment and route planning.

Tomas Larsson, Jeppesen

Airlines operate in a complex environment. We cannot absorb all complexity as it is, so we find ways to reduce complexity. We divide our own operation into expert areas, and we aggregate the world around us into groups. Optimization is a powerful tool to master complexity, but the optimizers are designed to master one expert area, like scheduling or crew planning. By extending the competence of the optimizers to have some overlap with adjacent areas, they can accept greater complexity. This presentation provides an example from integration between scheduling and crew planning, where we quantify the value generated when the capacity of scheduling to increase its capacity to absorb crew complexity is increased.

Pranav Gupta and Renzo Vaccari, Amadeus

Airline scheduling is a complex task that involves the efficient assignment of fleet and placement of flights at optimal times while considering various operational constraints. Schedule optimization primarily focuses on operational efficiency and profitability, and the schedule is handed over to crew planning teams. Crew scheduling involves creating efficient crew duties and pairings such that every flight is covered, and the crew must return to its base. This siloed process is suboptimal and solving an integrated problem has been a research challenge. Recent research and industry trends have highlighted the importance of crew-friendly considerations during schedule optimization. We discuss some important tenants of crew-friendliness in a multi-objective optimization framework that incorporates various crew preferences, such as rest time, duty time, and layover duration, in a manner that does not increase the complexity of the scheduling problem. To evaluate the effectiveness of our approach, we conduct simulations using major airline data in North America. The results demonstrate that our approach improves crew productivity and minimizes crew costs while maintaining profitability.

Suki S. Ng and Derek Chau, Cathay Pacific Airways

Schedule planning is a composite of science and experience, where operating strategies, business assumptions and business outlooks are often converted into various levels of planning assumptions and cascaded to multitude of short, medium and long term planning. Looking from execution perspective, there are two weak links in this process, namely operational dynamics and change in risk perception, requiring airlines to frequently tune its policy implementation and yield to suboptimal policy design. In Cathay Pacific, we have adopted an optimization driven simulation strategy to allow quick experimentation of scheduling policies in digital environment. Assumption validity, decision effectiveness, risk exposures and cost implications in operational environments are thoroughly tested and evaluated, enabling a systematic evaluation and profiling on policy parameters, providing a comprehensive capability to plan and experiment new way of work in the post-pandemic era.

Ronald Chu and Hossein Dashti, American Airlines

It is well known that the objective of a profitable schedule and an operationally reliable one often works against each other, and it has been a long challenge for the OR practitioner to design tools to find a balance for both. In this talk, we discuss some recent successful enhancements to our scheduling tools that improve both objectives. We explain how we implement the Demand Driven Dispatch model without compromising operation integrity while capturing additional bookings and revenue opportunities. We will also discuss how we insert an Equipment-with-Crew process to bring crew pairing and aircraft routing together to enhance operation reliability while creating significant cost saving. While we are still far from a fully integrated model of fleet assignment, crew pairing, and aircraft routing, it is a small practical step towards the laurels everyone is working for.

Markus Kühlen and Klaus Lütjens, German Aerospace Center (DLR)

We present a mixed-integer optimization approach to model the fleet assignment in the global air transportation system. Within the objective function of the optimization, cost and revenue aspects are covered by modeling the direct operating costs and the perceived attractiveness of transport offerings from a passenger’s point of view. The optimization constraints consider, among others, airport capacities, aircraft payload–range performances, and available take-off and landing field lengths. We calibrate the global fleet assignment model with historic fleet and flight schedule data of the period 2010-2019. Finally, we apply the model to analyze the attractiveness and potential operating scenarios of future narrow-body aircraft designs with different climate-friendly propulsion architectures (power-to-liquid, liquid hydrogen and hybrid-electric).

Mohand AIT ALAMARA, Air France

Given a flight schedule and a set of aircraft configuration, the fleet assignment purpose is to determine which type of aircraft should fly each scheduled flight. The proposed model aims to maximize the total revenue depending on flight features (duration, frequency , passenger unconstrained demand) and aircraft type characteristics (fleet size, cabine capacity, operational costs , freight revenue , preferential assignments). The passenger revenue computation is based on the spill model. It computes an estimated passenger traffic on a route / cabine level, given an unconstrained demand , a seat capacity and a demand distribution parameter. The non-linear shape of the spill function makes the assignment problem very hard to compute. We propose a MIP formulation of the fleet assignment problem that overcomes the non-linear spill complexity. In addition , we prensent an entension of this model in order to adjust the schedule to a fleet size evolution.

Reza Baharnemati, Amadeus

Kaori Bray, Southwest Airlines

Airline network forecasting is a crucial process for airlines to optimize their network planning and strategic decisions. However, forecasting accuracy is often affected by various factors such as demand uncertainty, market dynamics, operational disruptions, and external shocks. We propose a framework to measure the sensitivity of optimal schedules to changes in forecasted demand and revenue. We use historical data from a major US airline to compare the forecasted and actual network performance indicators such as passenger demand, revenue, load factor, and profitability. We also illustrate how the airline could influence different levels of forecasting accuracy, the optimal network design, and strategic choices of the airline by incorporating additional data sources in the forecasting process. We find that forecasting accuracy has a significant effect on the airline's network planning and strategic decisions and that improving forecasting accuracy can lead to substantial benefits for the airline. We also discuss the implications of our findings for airline network forecasting and planning practice and suggest directions for future research.

John Pepper and Josephine Dietrich - Allegiant Travel Company

Our presentation would cover the current state of pilot supply challenges in the United States, provide our forecast for both pilot demand and pilot supply, provide context/data on the drivers of pilot supply / demand, and show the impact that the lack of pilot supply is having on both domestic and international recovery from COVID. We would also present proposed solutions for the various bottlenecks in the pilot supply pipeline. While our research is most applicable for US market the ramifications are being felt around the world.

Pranav Gupta and Renzo Vaccari, Amadeus

Every year, the global airline industry faces millions of dollars in cancellation costs, with crew legality and curfew violations being two common causes of flight cancellations. These cancellations can be minimized through robust scheduling practices. Crew legality refers to the requirement that airline crew members must adhere to duty limits, and duty limit violations can result in flight cancellations or the need for replacement crew members. Similarly, curfew violations occur when flights operate during restricted hours, leading to diversions, cancellations, or fines from airport authorities. To address these risks, we propose a data-driven simulation-based approach that estimates crew legality risk and curfew violation risk in a schedule, followed by a MIP-based optimization model to minimize these risks. Proactive duty splits can mitigate crew risk while mitigating propagating delays can minimize curfew risk. To validate our approach, we conducted simulations using historical data from a major European airline, demonstrating that our approach can minimize the most significant delays.