56th Symposium Technical Program

The Role of Human Intervention with RM Systems. User Influence: What is it good for?

Bill Brunger (AGIFORS Fellow)


The Airline Industry has invested heavily in both systems and people to improve Revenue Management. The impact of system developments has been well documented – Less attention has been paid to the evolving role and impact of human/analyst intervention in the improvement of revenue results. Using the PODS simulator, supplemented by experience running and advising RM Departments, the author will examine User Influence in the context of RM System development.


Data-Driven Models for Itinerary Preferences of Air Travelers

Rodrigo ACUNA-AGOST (Amadeus)


There is an increasing interest within the travel industry in better understanding customer behaviour, particularly the way customers choose between a defined set of flight alternatives when searching for flights. Such an understanding can help travel providers (e.g., airlines) adapt better to market conditions and customer needs, thus increasing their revenue. In this presentation, we describe a two-stage approach to predict traveller choice behaviour by combining machine learning and discrete choice modelling techniques. The applicability of the models is illustrated by an application for dynamic pricing optimization. We conduct experiments on a dataset extracted from searches and bookings on several European markets. The experiments show that the proposed models have good levels of accuracy, permitting the dynamic pricing engine to obtain a significant increase of shopping session contribution, ranging between 60% and 80%.


Real-time Optimisation of Operational Slot Assignment

Valentin Weber (Amadeus) and Jeremy Baillie (Qantas)


To manage airport congestion, Air Navigation Service Providers (ANSP) implement Ground Delay Programs (GDP) at their major airports. Airport capacity is managed through slots, which can be assigned or swapped between flights and airlines through a Collaborative Decision Making (CDM) approach. Despite the variations in the implementation of CDM from one ANSP to the other, airlines must operate their flights in compliance with their assigned slots.

Slot management is mostly done manually by dedicated operators. To support them, we propose a continuous optimisation approach that automatically reassigns slots in order to minimise the associated ATFM (Air Traffic Flow Management) delay. This approach has been implemented in a production decision support tool. We will present how this live application has been successfully deployed inside an Operations Control Center and what are now the benefits for the airline.


Preference-Based Trip Trade System for Pilots

June Ma (American Airlines) *Crew Study Group Best Presentation*


We developed a preference-based trip trade system for American Airlines pilots to enhance their schedules. The system allows pilots to drop sequences (pairings) to, pick up sequences from, and trade sequences with other pilots and a pool of sequences that are currently not owned by anyone. It honors pilot seniority with fairness, FAA mandates, and company rules using a comprehensive modeling framework. We will briefly describe the system and share implementation challenges, some examples, and lessons learned.


Airline Maintenance Basing Strategy & Inventory Planning - The Right Parts, Right Place, Right Risk, Right Cost

Dawen Nozdryn-Plotnicki (Boeing)


Boeing Business Consulting and the Analytics team work with airlines to first provide a new maintenance basing strategy, and then to centralize their maintenance operations inventory. The line maintenance inventory spare part allocation is projected to save tens of $M. Some key questions answered were: Where and when to do maintenance – visual & descriptive analysis; How much line maintenance inventory and where to allocate them in the network – simulation based optimization


Predictive Maintenance Check from Analysis of Airplane Sensor Data

Dawen Nozdryn-Plotnicki & David Kinney (Boeing)


Unscheduled maintenance drives 10% of the annual operational cost to airlines worldwide. Predictive Maintenance could reduce those costs, particularly when synchronized with airline’s operations. By using engineering expertise, statistics and machine learning on aircraft sensor and fault data, as well as analysis of an airline’s flight and maintenance schedule, we detect impending issues on the aircraft and suggest maintenance tasks in accordance with the prediction and an airline’s working rhythm. These predictive maintenance tasks will increase reliability and reduce unscheduled maintenance.


Airline Operations in Digital Airline - Part I

Lukasz Stankiewicz (Boeing)


Airlines are moving to digital world more and more. Some of them are there, some just starting. In order to help airlines better understand benefits and challenges of integration and automation, we created the Virtual Digital Airline (VDA). VDA is simulated environment of existing Jeppesen/Boeing products and tools to manage personnel, assets, and workflows. This presentation will demonstrate a novel approach to conduct research on future airline operations by using this simulated environment. Developed approach enables collection and analysis of customer feedbacks and address enormous challenges airlines are facing today.


Aircraft Boarding Strategies

Massoud Bazargan (Embry - Riddle Aeronautical University)


Airlines start generating revenues while their aircraft are flying. Reducing aircraft turn-around times is an important goal with passenger boarding being a major metric. A simulation approach to aircraft boarding strategy was presented at AGIFORS Airline Operations conference a few years ago. The study considered a single aisle narrow body aircraft such as B 737 or A-320, where passengers board through the single front jet way. The objective of that study was to identify key performance measures such as total boarding times and number of seat and aisles interferences for popular boarding strategies such as back-to-front, window-middle-aisle and random. An optimization model was also proposed and its performance compared to other strategies was demonstrated. As the airlines started charging for bags, it appears that the boarding times are taking longer, independent of what the strategy they implement. In this study, a new technology is proposed for the above single aisle aircraft jet way, which attaches to the existing facility and enables the passengers to board through both front and back of the aircraft. The objective of this revised strategy, is to revisit the KPIs and evaluate the boarding strategies, through one and two jet ways. In addition, this study considers the deplaning times with one and two-doors. Deplaning has not received much attention and is an important factor for airlines implementing hub and spoke to insure the passengers can make the connections times. The results are very encouraging for the airlines in general and low cost carriers in particular. The study may potentially motivate the airlines and/or airports to seriously consider adopting this technology.


Cabin Configuration Analysis

Himanshu Jain and Mark Diamond (ICF)


Cabin configuration analysis of an airline’s fleet can be a complex task, which is affected by many factors such as different aircraft types in the fleet, cabin products, evolving network and fleet plans, passenger loads and yields by product, market growth and other issues. There are important tradeoffs and opportunity costs associated with the space occupied by every seat on the aircraft. In this presentation, we will describe a case study in which ICF helped a global hub-and-spoke airline identify more than $80 million in added contribution per year from reconfiguring several aircraft types in its fleet. We will talk through the situation, challenges, and the data-driven approach which we used to analyze the value generated by each cabin product and to come up with optimal cabin configurations.


An Integrated Optimization Approach to the Challenge of Aircraft Maintenance Planning

Luis Alvarez (INFORM)


As more and more aircraft take to the skies and engineering skills remain in short supply, the future of aircraft maintenance will require innovative ideas to continue carrying out maintenance efficiently and safely. Tail assignment, maintenance staff planning and work package planning play central roles in how overall maintenance is carried out. Although each of these depend on each other, optimization solutions have so far looked at them only individually.

Advances in hardware and optimization algorithms open up new opportunities to start taking an integrated optimization angle involving all three areas and offering benefits not possible until now. In this talk we will explore the challenges associated with taking an integrated view, discuss the possible solutions and present some first results showing the promises of this approach.


On The Move towards Full Crew and Fleet Integration

Tomas Gustafsson (Jeppesen)


The traditional planning steps within different airline resource areas inherently lead to various levels of sub optimization. Crew planning is typically more complex and more constrained than aircraft allocation, yet general practice is to let the aircraft rotations be input to crew planning. Here we discuss how that process can be turned around by allowing crew planning to influence aircraft rotations already from an early phase. We'll compare scenarios for a simple iterative approach, with higher levels of integration together with the option of also allowing re-fleeting. Computational results from multiple airlines will be shared.


Optimizing Baggage Handling Using Lean Six Sigma

Willem-Jozef Van Goethem (Kenya Airways)


This case study is an implementation of the Lean Six Sigma (LSS) method to improve on-time performance and in specific reduce baggage related delays at Kenya Airways. The project contributes to the financial performance as delays impact ground handling, passenger compensation and so-called “soft” costs or lost revenue opportunities. The root causes for process variation were identified within three main areas: out-station loading sequence, baggage off-loading and baggage loading. Three pilot cases studies are currently being executed to test solutions to improve loading sequence, redesign the narrow body turnaround process and improve team work between baggage handling and ramp teams by aligning KPI’s. Using the LSS improvement approach, Kenya-Airways has redesigned its baggage handling process to be better in line with it hub and spoke business model. Early estimations indicate that improving the baggage handling process can save 7 % delay-related costs.


Digitalization in Flight Operations and the Impact on Operations Research

Jeffery Oboy (M2P Consulting)


Technology has transformed how businesses and consumers connect, sell and work with one another in virtually every industry around the world. Be it through mobile, internet of things, cloud computing, big data, analytics or social media, the impact has not only been massive, but it will grow even further in the next few years. While the airline industry has concentrated the majority of its digitization efforts on commercial aspects thus far, focus is now shifting more and more so to operations. The question is what the impact will be in operations and what this could especially mean for the field of airline operations research.


A Multistage Stochastic Programming Model for Air Cargo Assignment under Capacity Uncertainty

Felipe Delgado (Pontificia Universidad Católica de Chile)


Air cargo transportation represents an important source of income for airlines. Cargo can be transported by two means: with freight planes, or in the belly of the passengers aircrafts, a practice known as "belly cargo". One of the main difficulties faced by the belly cargo mode is the uncertainty on the capacity, due to the fact that the actual availability for cargo is only known after the passengers have boarded, and had their luggage loaded. We propose a multistage stochastic programming model whose goal is to find the optimal allocation of cargo to the passengers network in order to maximize profit, taking into account incomes, costs and penalties for not delivering cargo that was previously accepted. The resulting problem is a large scale linear program, and we use decomposition techniques in order to be able to find good quality solutions and bounds for the optimal value. Our numerical experiments are based on a real network of a major commercial airline.



Nonparametric Estimation of Customer Segments in Airline Revenue Management

Johannes Ferdinand Jörg (RWTH Aachen University)


Customer segmentation is a crucial prerequisite for optimising the firm's interface to the customer, such as through revenue management and pricing. Today, technology lets firms access increasingly big data sets. This creates the challenge to extract the information contained in such data to incorporate it in decision making. This presentation shows a nonparametric approach to estimate customer segments from airline booking panel data. By considering fare class availability data, we can employ the approach to censored panel data. On several synthetic data samples, we validate estimation performance in a controlled environment. On empirical airline panel data, we demonstrate the practical applicability of the approach.


Robust Revenue Opportunity Modeling

Richard Ratliff – Sabre


This presentation describes research work in progress on a new, QP-based revenue opportunity model intended to optimize revenue while providing market-level allocations which are more stable and robust over time than traditional, LP-based ones. Because airline O&D networks foster passenger connections, it results in more markets served than flights operated; this structure provides additional degrees of freedom for RM bid-price control solutions with alternate optima (or near optimal) revenue. Although these different alternate solutions can lead to the same (or nearly the same) network revenue outcome, they cause manageability issues for airline RM analysts in practice. A desirable feature of a ROM solution is, to the extent possible, to generate similar types of market-level controls over time (e.g. in a market such as JFK-FRA on Tuesdays, keep the local traffic closed and the flow markets open). Such stability aids RM analysts in setting effective default allocations and monitoring outliers; it is also a consideration when holding RM analysts accountable to market-level ROM performance metrics.


Critical Assessment of Five Methods to Correct For Endogeneity in Discrete-Choice Models

C. Angelo Guevara (Universidad de Chile)


Endogeneity often arises in discrete-choice models, precluding the consistent estimation of the model parameters, but it is habitually neglected in many practical applications of airline demand, invalidating pricing and revenue-management analysis in this context. The purpose of this presentation is to contribute in closing that gap by assessing five methods to address endogeneity in this context: the use of Proxys (PR); the two steps Control-Function (CF) method; the simultaneous estimation of the CF method via Maximum-Likelihood (ML); the Multiple Indicator Solution (MIS); and the integration of Latent-Variables (LV). The assessment is first made qualitatively, in terms of the formulation, normalization and data needs of each method. Then, the evaluation is made quantitatively, by means of a Monte Carlo experiment to study the finite sample properties under a unified data generation process, and to analyze the impact of common flaws.


Some Recent Improvements in Operations Recovery Management

Xiaodong Luo (Sabre Airline Solutions)


In this talk, We are going to present some algorithmic and modeling improvements we did in the operations recovery area in the past few years. Specifically, we will cover a smart flight delay generation procedure, a sequence LP approach for integer multi-commodity network flow schedule recovery problems

as well as a new optimization model for effectively handling connection based minimum ground time. We will present computational results to demonstrate the effectiveness of these models and techniques.


Next Gen Ops System Implementation at Qatar Airways

Piyush Taori & V.P. Praveen (Qatar Airways)


Qatar Airways recently implemented an in-house developed Ops Control System, TOPS (Total Operations System). The new system covers whole spectrum of business processes of Aircrafts & Flights, right from long term maintenance planning to short term maintenance planning and then towards the day of ops starting from the tail assignment to the day of ops movement tracking and an extremely sophisticated flight watch module compliant with the upcoming IATA 4D/15 requirements. The system is highly integrated, uses OR based tools for complex optimization problems (e.g. Tail Assignment and Maintenance Planning) and local solvers for small scale problems which need immediate decision support. The system has a real time rules and violations module which alerts the user in case of any exceptions. An extremely feature rich ""workspace"" module allows users to carry out what-if analysis.

The road map of the system includes (development started) the crew control module.


Anna Valicek Competition Papers


The two finalists for the Anna Valicek Award will also present their papers. 


Metaheuristics for Efficient Aircraft Scheduling and Re-routing at Busy Terminal Control Areas

Marcella Sama (Roma Tre University)


Customized Offers in Airline Revenue Management

Michael Wittman (Massachusetts Institute of Technology)


Contact us

Address: AGIFORS President, Georgia Institute of Technology, 790 Atlantic Drive, Atlanta, GA 30332

Phone: +1 (404) 385-6634

Email: president@agifors.org

Powered by Wild Apricot Membership Software