AGIFORS Annual Symposium 2003
Technical Program

Keynote Speeches

Ethnography and OR
Bill Brunger, Continental Airlines

The speaker will report on his own Doctoral research contrasting an OR approach with an ethnographic approach to analyzing a customer's arrival at an airport.

How can software and change become more compatible, and how can companies better leverage their optimization and OR experts?
Pierre Haren, ILOG 

In the past 2 years, after the Y2K and euro upheavals, the main driver for new IT projects has been the need to adapt corporate software to ever-more rapidly changing business conditions: new competition (such as low-cost carriers for airlines), new regulations (such as the 35h week in France, the Sarbanes-Oxley act in the US), new catastrophic events (9/11, SARS). The list goes on.  Applications with infrequent releases and that require the intervention of OR experts to cope with changes are increasingly hampering the ability of a corporation to successfully cope and compete. To quote a European CIO "I'd rather have a good solution to today's problems than an optimal one to yesterdays." Recently, the finance industry has adopted "Business Rules" to increase their agility and responsiveness.  This represents a new paradigm for IT.  Can we derive some lessons from this success, and apply the same concepts in order to envision an increased influence of optimization and OR experts in corporations?

Presenting to Your Senior Executives: O.R. That Even a V.P. Can Understand
Scott Nason, American Airlines

 

Task Assignment Optimization Model
Barry Smith, Sabre Holdings

Airport staffing is a significant expense for airlines and ground service providers.  Sabre developed a staff scheduling process that involves task identification from flight schedules, construction of shifts and assignment of tasks to shifts.  Assignment of tasks to shifts is complicated by task qualifications and travel time between tasks.  We developed a task assignment optimization model based on graph theory and integer programming.  The model produces significantly better results compared to the heuristic it replaced.    

Presentations ( Based on submitted abstracts to date* )

A Multi-Criteria Optimization Approach for Robust Aircraft Rotations
Irina Ioachim, Delta Technologies

Building robust aircraft rotations is a challenging task due to the multi-criteria optimization nature of the problem. Moreover, the problem presents a very complex combination of soft and hard constraints. In this talk we will present an approach which combines the use of daily, weekly and fully-dated models with global and heuristic optimization techniques.

Benefits and Costs of O&D Revenue Management
Rick Zeni, USAirways

US Airways has been using true bid price O&D control for the past four years. During this presentation we will share our experiences with the benefits and challenges faced by the revenue management group at the airline. We will review the evolution of the business process as well as system implementations.

Choice-Based Revenue Management: Research Advances and Implementation Prospects
Garrett van Ryzin, Columbia University

Using consumer-choice models as a basis for revenue management (RM) is  appealing on many levels. Choice models can naturally model important buy-up and diversion phenomenon and can easily integrate newer, undifferentiated low-fare structures. Despite its vast potential, choice-based RM is a radical departure from current practice and the task of revamping RM technology to incorporate choice models has been a daunting prospect. Yet recent research advances have now brought choice-based RM within striking distance of being practical. In this talk, we survey this research and discuss its implications for RM practice.

Codeshare Optimizer – Maximizing Codeshare Revenues
Raj Sivakumar, United Airlines

Code share agreements allow airlines to extend the reach of their networks by placing their code on the other carrier’s flights. The airline industry is currently seeing a surge of code share agreements, partly as a means to combat the adverse revenue situation in the industry. Some of the key examples are the United Airlines-US Airways code share and the three way code share between Delta, Northwest and Continental. The conventional wisdom on code share has been that greater the extent of code share, in terms of flights and connections, more is the revenue upside to each carrier. We discuss the economics of code share. We also present the Code Share Optimizer, a network optimization tool built by United’s Research and Development group that considers the complex interaction between the demand, fares, market shares and the prorate agreements to recommend optimal code share levels.

Decision Support for Hub Reduction
Tim Niznik, American Airlines

This presentation discusses the development and implementation of an automated tool that identifies alternative cancellation plans to address operational irregularities such as those caused by snowstorms, thunderstorms, and ATC delays. These plans are constructed so as to minimize the impact on passengers, crew, and aircraft routings using an approach that combines subject matter expertise with mathematical optimization techniques.

Large Scale Crew Rostering
Stefan E Karisch, Carmen Systems AB

We first give an overview of the Carmen Crew Rostering system, which, until recently, could efficiently solve crew rostering problems with up to 3000 crew members. Then, we present enhancements to the standard approach used in the production system that makes it possible to solve large scale rostering problems of up to 6500 crew to production quality in less than 14 hours.

Low Cost Carrier a new area of RM
Ruediger Thiel, Lufthansa Systems

Revenue Management was created as a weapon against Low Cost Airlines in the deregulation price wars during the 80th., which was won by the established airlines The recent focus was on full network optimization and O&D controls. With the resurrection of Low Cost, as “the” successful airline business model of the new millennium, Revenue Management faces a new challenge. The presentation focuses on the impacts on revenue management and describes a new approach for revenue management considering the business rules of the low cost airlines model.

Managing Innovation at Sabre Labs
Richard Ratliff, Sabre Holdings

Given the major changes recently in the travel industry, innovation is becoming increasingly important for companies to successfully adapt to the new business and technology environment. This presentation provides an overview of innovation management techniques and common problems encountered in practice. Specific examples from Sabre Labs are discussed.

MultiFlight AU Analyzer
Judy Pastor, Continental Airlines

Most no show forecasting and subsequent AU (authorization level) setting is done on a flight-by-flight basis. This myopic approach ignores the effect of oversales and subsequent rollover of passengers of previous flights in the market. The onus is on the RM Analyst to take this effect into account outside the modeling system. This tool, developed in-house, is a first step to allow analysts to examine their AU decisions on the entire market via a simulation. Some results and topics for future research will be covered.

Probabilisation of Demand Forecasting
Daniel Sallier, Paris Airports Authority (ADP)

Most of the strategic projects in the air transportation industry require a demand forecast to be performed first. The forecasters come up generally with either a single demand figure or a set of figures related to different scenarios (base case, worst case, high case). By doing so, the forecasting team set implicitly the level of financial, economical and industrial risk his company might be faced with. ADP's R&D Team has developed the methodologies and the techniques to produce probabilised demand forecast opening the path for more rational and quantitative and less subjective assessment of the risks attached to any project.

Quantifying Resource Impact on Schedule Profitablity
Tim Jacobs, American Airlines

Traditional airline fleeting models focus on the efficient allocation of scarce aircraft resources while maximizing overall profitability of the schedule. These models rarely incorporate the other limiting resources such as flight deck, cabin and ground crew availability. This paper presents an approach to quantifying the impact personnel resources have on schedule profitability. Results indicate that limited overall profitability increase can be achieved as support resources are reduced due to improvements in crew efficiency. As the reduction in support resources continues, adverse impacts begin to emerge on the overall profitability by forcing the cancellation of profitable flights. Our paper shows how airlines can establish optimal support resource levels, at which overall profitability is maximized.

RM overview: challenges and opportunities
Umit Cholak, Sabre Airline Solutions

This presentation will cover a brief history of Revenue Management in the airline industry, and it's current and future challenges. It will try to strike the right balance between theory and applications and attempt to bridge the gap between RM and Reservation systems. The presentation will also attempt to identify areas of opportunities in the future.

Revenue Management Product Design
Kalyan Talluri, Universitat Pompeu Fabra

Designing appropriate revenue management products has been a neglected topic in RM research and systems design. With new low-cost entrants changing the competitive landscape dramatically with new products, the issue has come to fore for many established carriers. We present a model that, based on transactional data, designs a set of nested products for sale in an RM system, along with associated prices and aggregate allocation recommendations.

Revenue Management with Restriction Free Pricing
Shankar Mishra, Sabre Airline Solutions

In order to address lower traffic volumes and increased price sensitivity of travelers, airlines have started offering restriction-free fares for all or parts of an airline network. This eliminates any round-trip purchase, Saturday night stay, blackout period requirements, etc., on fares and empowers revenue management with the complete inventory control responsibilities. In this presentation we discuss the impact of such a pricing structure on inventory controls and explore revenue management solutions to support that.

Robustness Issues in Pairings Optimisation
David M. Ryan and Matthias Ehrgott, University of Auckland

Besides constructing aircrew Pairings with minimal cost, airlines also wish to construct pairings which are robust in the sense that flight schedule disruptions are less likely to propagate delays into the future. In general a minimal cost solution is likely to lack robustness and conversely a solution with maximum robustness (however this is to be measured) is likely to be more expensive. A measure of robustness for each pairing will be developed and the concept of a robustness objective will be discussed. The two objectives of cost and robustness will be treated in a bicriteria optimisation to generate "efficient" pairings which do not allow a simultaneous improvement in cost and robustness.

Seeding Up Column Generation Via Convex Optimization Tools: 
An Application to the Aircraft Rotation Problem

Olivier du Merle, Air France

Proximal ACCPM is a new method for nonsmooth optimization. This method is a variant of the analytic center cutting plane method (ACCPM), in which a proximal term is added to the barrier function that defines the center. In this talk we will give a presentation of the method and of its implementation. Numerical results concerning real aircraft rotation problems are given comparing Proximal ACCPM and more classical column generation approaches.

SimDelta: A Sim. Model to Analyze Airline Ops
Wassim Chaar, Delta Airlines

Airline operations are subject to many uncertainties and anecdotal evidence suggests that most airlines never experience a day without disruptions. Disruptions vary in size and effect, resulting in anything from minor delays to major cancellations. For large airlines, such as Delta, the gap between planned costs and actual operating costs can be in the range of several million dollars. Understanding the inherent sources of operational uncertainty in order to develop operationally "robust" plans, can be an extremely daunting task. We present a large scale simulation model to analyze airline operations and discuss results of some basic scenarios to understand schedule performance.

The cabin crew rostering model at Alitalia
Francesco Sansone, Alitalia Airlines

The Alitalia crew monthly rostering process for the cabin crew members is, nowadays, supported by an automatic system called Saturn. This system is very efficient and produces high quality rosters but is a hardly manageable by user model too.  Alitalia OR department is reengineering this system in order to have a modularised system in which the model is separated from the Agreement and Quality (i.e. even distribution, crew preferences) rules. The project takes advantage of Carmen Consulting cooperation, in fact the modelling language Rave constitutes the interface between the model and the rule modules. The model uses Rave like an oracle asking for legality checks and quality information in order to satisfy the constraints and to optimise the objective function. The assigning algorithm is based on a generalisation of the Hungarian algorithm.  This presentation will concern the architecture of the new model shell.

The Evolution of Airline Yield Management: What YM Capability Does My Airline Need?
Ben Vinod, Navona Inc.

The evolution of yield management began with the deregulation of the airline industry in 1978. Airline yield management capabilities have advanced from leg/segment class controls to virtual nesting and continuous nesting controls. The increasing prominence of low cost carriers have given birth to a new genre of yield management based on optimal pricing along the demand curve to shape demand. With the available options, an important question faced by every airline in the world that operates in a competitive environment is the investment and the precise capability needed to deploy a successful yield management program. This paper provides guidelines on determining the yield management capability required for an airline to generate the incremental revenues.

The Repair-Game: Robust Plans and Disturbation Management in Aircraft Scheduling by the Help of Game Tree Search
Ulf Lorenz et. al., Paderborn University

Subject of our research are plans of logistic processes, especially aircraft schedules from airline companies.  In this project, we want to combine game tree search algorithms --- as we know them from games like chess --- with conventional optimization algorithms of Operations Research, in order to enable the operation managers to quickly and efficiently react to schedule disruptions. The aim is to find back to the original plan, as close as possible. Here, we do not only want to evaluate an is-situation by inspection of the current state of the system, but we also want to take into consideration further possible events resp. future scenarios. Our methods shall be able to support planning teams in finding robust plans. I.e. such ones, that can easily be lead into the original plan.

Call for papers

The 43rd Annual Symposium of the Airline Group of the International Federation of Operational Research Societies (AGIFORS) features a comprehensive program covering the whole spectrum of airline operations research. This year’s symposium will be held from September 14 – September 19 in Paris, France.  Contributions to the technical program are solicited in the following areas, but are not limited to, flight scheduling, pricing, yield management, maintenance and engineering, crew, ground resources, and finance.  Papers covered by active Study Groups, as well as other topics not covered by active Study Groups, are sought.  Airline, consultant, academic, and industrial research representatives are all encouraged to attend.

Presentations are 30-35 minutes, with 10-15 minutes for Q&A.

Symposium proceedings will be published by January 2004.

Deadlines and Requirements

Paper Submissions

Please send all submissions via the online form by 15 August 2003.

Paper Inquiries

Thomas Fiig, M.Sc., Ph.D.
AGIFORS Symposium Technical Program Chair
Scandinavian Airlines System
Hedegårdsvej 88
DK-2300 København S.
Denmark
Phone: +45 3232 2190
E-mail: Thomas.Fiig@SAS.dk

August 15, 2003

Submit abstracts online (maximum 100 words). Include a title and identify the corresponding author.  For each co-author, include full name, affiliation, complete address, telephone number, FAX number, and electronic mail address or URL.

September 1, 2003

Corresponding author provides both an electronic file submission of the presentation/paper for inclusion on the CD-version of the Proceedings and a photo-ready copy for the printed Proceedings.

Please note the following conditions for inclusion of your paper in AGIFORS 2003:

  • At least one author of an accepted paper must attend the Symposium to make the oral presentation.
  • Each attendee must register and pay the registration fee.
  • Each attendee is responsible for making all hotel arrangements.

Presentations ( Based on submitted abstracts to date* )

Title

Speaker - Affiliation

Abstract

Panel Discussions

Title

Panelists

  • (moderator)

Vendor Presentations

Company listing

* Technical Program is subject to change

Welcome | Agenda | Registration | General Information | Hotel Information | Host City