|
AGIFORS Annual Symposium 2008 |
||||||||||
|
Technical Program (Based on submitted abstracts to date*) These submitted abstracts have been conditionally accepted for inclusion in the AGIFORS Annual Symposium 2008 conference agenda. Final acceptance is contingent on 1) receipt of complete presentation on/before 07 September 2008, and 2) approval of presentation content by the AGIFORS technical committee. Monday - 22 September 2008 Web Services and Operations Research To what degree can Operations Research models be plugged together using web services? This presentation will briefly cover web services standards and Sabre's experience in implemented them for reservations and online operations, with more than 1200 customers, including 34 airlines. I'd like to initiate a discussion on web servicess for decision support applications and AGIFORS role in creating open standards for our industry. The goal is to make it easier for airlines to install systems, to use systems and components as part of their own processes, and to exchange data with our academic research partners. Modeling the Choice of an Airline
Itinerary and Fare Product In this research, we develop a methodology to analyze the choice of an airline itinerary and fare product based on booking data. Since non-booked travel alternatives are not recorded in airline bookings, booking data was combined with fare rules and seat availability data to incorporate the impact of airline pricing and revenue management and reconstitute the choice set of each booking. In addition, characteristics of the traveler and the trip were retrieved from the booking records and replaced trip purpose that is traditionally used to segment demand in airline markets. They were included as explanatory variables of a latent class choice model in which several factors can be used simultaneously to segment the market without necessarily dividing the bookings into many small sub-segments. In addition to an improvement in fit over previous models based on a deterministic segmentation of the market, the latent class choice model was found to provide a more intuitive segmentation of the demand between a core of time-sensitive business travelers and a mixed class of price-conscious leisure and business travelers. This model extends the range of applications of passenger choice models to additional airline planning decisions such as pricing and revenue management. Decreasing Airline Delay Propagation
By Re-Allocating Scheduled Slack Passenger airline delays have received increasing attention over the past several years as airspace congestion, severe weather, mechanical problems, and other sources cause substantial disruptions to a planned flight schedule. Adding to this challenge is the fact that each flight delay can propagate to disrupt subsequent downstream flights that await the delayed flights’ aircraft and crew. This potential for delays to propagate is exacerbated by a fundamental conflict: slack in the planned schedule is often viewed as undesirable, as it implies missed opportunities to utilize costly perishable resources, whereas slack is critical in operations as a means for absorbing disruption. In this paper, we show how delay propagation can be reduced by redistributing existing slack in the planning process, making minor modifications to the flight schedule while leaving the original fleeting and crew scheduling decisions unchanged. We present computational results based on data from a major U.S. carrier, showing that significant improvements in operational performance can be achieved without increasing planned costs. CitationShares from the Lab to the
Bottom Line CitationShares operates about 90 business jets, with guaranteed availability of an aircraft for customers who call with at least 8 hours' notice. Last-minute changes require us to reschedule within hours of departure more than ten percent of the time. To help us cope with the challenging combinatorial problems we have to solve on the fly every day, we have implemented the Scheduling Optimizer. Schedule costs have fallen from over 20% above optimal to within 5% of optimal. I'll discuss how we developed this tool, the challenges we faced incorporating it into our processes, and where we go from here. Virtual Spares Many airlines schedule aircraft connections at their hubs on a first arrive-first depart basis. Others deviate from this discipline to correlate aircraft connections with passenger connections, or to equalize aircraft utilization. Recent studies in the field of robust scheduling have shown that more robust schedules can be developed by correlating the ground-times of aircraft with the expected delay "upstream" of connections. However, the resulting schedules are not necessarily amenable to the swapping of aircraft between maintenance routings, a tried and true strategy for dealing with the realized (as opposed to expected) disruptions. I present a novel idea for scheduling aircraft connections that leverages the beneficial features of both approaches to mitigating the effects of disruptions. By scheduling the first arrival from an origin where operations are very reliable, and scheduling that same aircraft as the last departure and to a destination where operations are also very reliable, we create a "virtual" spare aircraft that can be swapped with any aircraft that is delayed. A similar strategy can be employed with the second, third, and other early arrivals to create additional virtual spares, subject of course to marketing considerations. Tuesday - 23 September 2008 Marketing Very Light Jets in a
Turbulent Economy A major initial problem that must be overcome in marketing very light jets is estimating the demand for this new type of service. Will the market be weak with one or two players, or will there be a lot of competition? How can such a new service provider sustain itself in an uncertain market? To attempt to answer these questions, the author will apply Kotter's (2007) "Eight Step Process of Successful Change" to the marketing of very light jets. A Complexity Theory Algorithm for
Routing and Scheduling Very Light Jets The challenge is to find a minimum “cost” schedule and aircraft routing to satisfy on-demand requests for service in a very light jet (VLJ) air taxi market. The solution is a Computational Complexity Theory heuristic that incrementally builds an aircraft schedule through a sequential insertion of on-demand requests specified by departure and arrival time windows. The algorithm minimizes the “cost” of insertion using a non-linear objective function (weighted sum of the disutility to the customer of an insertion). Computation times for 23 aircraft and 480 service requests are fewer than 5 seconds. Revenue Management Optimization at
Air Canada At the 2005 meeting in South Africa, Air Canada and EXPRETIO jointly presented a new framework to address the pricing optimization problem in Revenue Management. As it is part of the second generation of O/D models, it works with choice models. We will quickly reintroduce the model followed by some results. Some interesting aspects came from the applicability of the model (pricing optimization and/or seat allocation optimization) and the preprocessed steps (for example the segmentation) required to do the optimization. We will conclude the presentation by discussing the different potential applications (tactical vs strategic). The Rise and Fall of Airline Revenue
Management Systems Airlines worldwide have invested heavily in revenue management systems, supported by OR models for overbooking, fare class mix and network optimization. With the “simplification” of fare structures since 2000, airline RM systems are no longer able to maximize revenue without substantial modifications. The current RM challenge is to develop new approaches for forecasting and optimization in less restricted fare structures that can differ among airlines and which change regularly. This presentation describes several new approaches developed in the MIT PODS Consortium research, and suggests how OR can help to make airline RM systems more effective in the new pricing environment. Airline Customers Searching and
Purchasing Behavior Markov models are use to analyze how individuals search for information about air ticket attributes and prices using data from an airline website that employs a unique strategy for selling tickets at low prices to travelers with a high degree of travel flexibility. Results show search intensity is highly dependent on the relative discount level and there is a high degree of correlation observed across screens, that is, there is a “stickiness factor” in the sense that customers are seen to fixate on specific attributes of the search process. An Empirical Investigation of
Airline Flight Schedule Robustness Airline flight delays have come under increased scrutiny lately, with on-time performance at one of its worst levels in 2007. In the first part of this study, we combine flight data published by BTS, with the Newsvendor framework from the Operations literature to examine the impact of the scheduled block time on on-time arrival performance. Results show that airlines systematically under-schedule flights, and do not adjust service levels based on the time of the day, month, or origin/destination airport congestion. Next, we analyze the impact of flight block-times and aircraft turn-times on the schedule's robustness based on aircraft routings. Modeling the Impact of Eliminating
Service to Regional Airports As U.S. carriers continue to cut capacity in response to soaring fuel costs, one concern is the impact on small, regional airports. Many have seen a dramatic reduction in service and several have lost service altogether. What does it mean to eliminate service to a regional airport? From the carrier standpoint, the effects are observed not just on the eliminated flights but on the rest of the network as well. For the community, there are economic, environmental, and civic implications. As operations researchers, we can play an important role in modeling and analyzing these myriad and complex impacts. This talk will be a directed discussion of the factors that should be included in such analyses. Block Time Planning with Service
Levels Schedule planning divisions determine flight frequencies and preliminary departure times together with the corresponding fleet assignment. We provide concepts with models and a business case for computing block times based on the provided historical stochastic estimates. Operating cost and passenger revenue are captured with the latter at the itinerary level. Service level based on connecting traffic and the DOT on-time performance measure are key components. Proof of concept results based on data from a major US carrier will be demonstrated. Optimization Decision Manager Traditionally, organizations using optimization-based decision support tools have had to choose between relatively inflexible custom applications and continual modifications by Operations Research modelers. ILOG ODM bridges the gap between modelers and business stakeholders, facilitating the creation of flexible decision support applications with minimal requirements for IT support. ODM provides user-friendly GUI allowing business users to perform scenario-based “what-if” analyses, multi-objective goal programming and systematic constraint relaxation. ODM supports both Mathematical Programming and Constraint Programming optimization engines. We will present an overview of capabilities and a short demonstration of capabilities. Thursday - 25 September 2008
Integration of a Passenger Flow
Model into the Fleet Assignment Problem
Jonathan Dumas, Francois Soumis - GERAD The objective function of the standard fleet assignment problem (FAP) is linear in the FAP's binary decision variables. In consequence, it is hard to make it account correctly for the expected revenue associated with each tentative fleet assignment. We will briefly describe a passenger flow model that has nice properties (regarding network effects, spill and recapture, and the temporal and stochastic nature of the booking process), and then explain how we use it to iteratively improve the FAP's objective function. Airline Business Process Modeling In order to manage complexity of their operations, most airlines are organized in functional silos. Resulting fragmented processes proves suboptimal at best and can be costly, slow and inconsistent in operations. They can even be contradicting if all aspects of planning and operations are not consistent. In attempt to provide an efficient holistic solution, we have developed an analytical approach and created the airline business process model that maps identified functional areas and processes. Cross-functional processes are showing interrelated tasks at different levels of granularity. The developed model helps to measure efficiency, detect bottlenecks, create what-if scenarios and build robust, integrated mathematical models. Examples of crew planning and operation processes will be provided. A Heuristic Branch and Price
Approach to Repairing Crewmember Schedules Certain airlines within the industry have the daunting task of adapting their operations to many short-term disturbances that impact the flight schedule. We present a solution developed to repair crewmember schedules over a period of a few days, given new aircraft routes and a set of modifications made to the flight schedule. By focusing on operational constraints, the solution is driven towards simultaneously rebuilding crew routes and crew schedules. Cost and robustness objectives are balanced by imposing structural compatibility constraints between crewmembers that fly together. With results obtained in short runtimes, the solution can be seamlessly integrated into an airline’s daily processes. The Connectedness of the Worlds
Airline Network The impacts of service changes in one part of the airline network have unforeseen impacts elsewhere in the network. Using the global airline network representation from a worldwide agent-based model of passenger ticketing, it can be established that, under conditions of finite leg capacity and random-variable demand, it is impossible to consider any sub-network without the introduction of error. This presentation will discuss the magnitude and characteristics of that error, and show how the minimum error sub-network which encompasses any market set of interest can be defined, and will offer methods of estimating the expected error for such a sub-network. Are Internet Sales Less Profitable Customers who buy air transport through Online Travel Agencies find and purchase lower fares than through Traditional Agencies. The paper summarizes this impact, and then proposes a framework for comparing the relative profitability of different channels, controlling for customer and trip attributes. This is the final installment of the author's dissertation research, which has been presented at previous AGIFORS conferences. Restriction Free Pricing and Demand
Segmentation With the proliferation of low-cost carriers in recent years, an increasing number of markets are being priced with less restricted fare structures than were formerly in place. This has led to a violation of the assumption of independence of demand between fare classes, which is implicit in most commonly used forecast systems. This lack of demand segmentation coupled with today’s high fuel price has challenged the business models of many US carriers. We will discuss some of the approaches currently being applied at low cost and traditional network carriers. Multi-flight Demand Untruncation:
Simulation Results This talk provides an overview of a multi-flight demand untruncation approach with recapture using customer choice models. Results from applying this method in an O&D simulation are discussed. Untruncation (or uncensoring) involves estimating the true underlying demand from historical traffic and availability data. Many unanswered questions arising from new technical challenges in Revenue Management forecasting and optimization due to restriction-free pricing (RFP). Simulation is needed to properly consider the complex underlying interactions among different flights and fare classes in the market (including competitors). Maintenance Reachability Robust
Schedule and Route Construction at Southwest Airlines “Maintenance Reachability” – Robust Schedule and Route Construction at Southwest Airlines Irregular operations often ruin the best-laid aircraft maintenance plans, as aircraft bound for maintenance bases get re-routed for other reasons. At Southwest Airlines, this problem is further complicated by the need to balance maintenance work across multiple maintenance locations. This presentation will cover some planning techniques and metrics that Southwest Airlines is using to make the flight schedule more robust for handling both planned and unplanned maintenance events, as well as balancing work at the maintenance bases. How American Airlines Operations
Research team helps its maintenance operation The OR team at American Airlines (AA) has been collaborating with its Maintenance & Engineering business units since 2000 to improve the airline's overall maintenance operation. We have contributed both to building applications that help day to day operations and providing analyses that help M&E with strategic decisions. We worked with line, base maintenance as well as supporting organizations such as Inventory Management. Optimization, statistical analysis, and simulation among other OR/IE techniques are used in these projects. I will give an overview of a few key projects we have worked on and some challenges we have faced at AA. Friday - 26 September 2008 Evaluating Forecast Efficacy or How
Do I know if This Thing is Working There is much said and done regarding forecasting models. It is a challenge, however, to monitor forecasts, particularly in an O&D setting. In this presentation we will cover some of these challenges and investigate approaches for looking at and judging forecast efficacy. Upgrades and Upsells in Revenue
Management Capacity providers often experience a mismatch between supply and demand that can be partially alleviated while improving revenues by allowing for product upgrades. When prices are fixed and demands are independent, the problem is to decide which customer demands to upgrade to which products and when. We show that a fairness constraint can be imposed without loss of optimality under mild conditions. We also investigate a model that limits upgrades to the next higher quality product and provide necessary and sufficient conditions for its revenues to be as high as that of any less restricted upgrade model. Resellers of capacity also have an incentive to use upgrades as a mechanism to entice customers to higher quality products with higher commission margins. We show that this practice can be very profitable and that the profits can be much larger than direct commissions from sales would indicate. This suggests that primary providers may be in a more powerful position to negotiate commissions or impose sales volume constraints. We then investigate the case where sellers have pricing flexibility and customer demand is driven by a choice model. We derive pricing formulas under the assumption that demand for products follows a multinomial logit model. For this model we show that neither upgrades nor upsells can improve profits when margins are homogenous and there is complete freedom in selecting prices. However, upgrades can improve revenues significantly when margins are heterogeneous and sensible business constraints on prices are imposed. Yield Management: An Application of
Segmented Forecast The Yield Management has been apprehended as a technique of price modulation. Today, revisiting this vision seems to be necessary. The Yield Management which is a strategy of profit maximisation, of stabilisation and demand display cannot be reduced to a variant of price policy. Its application mainly draws on the interaction between the behaviour of the market and the optimisation of the income. It is in fact a consistent process which includes the management of the available units as well as the logic of prediction by customers' segments. This step should take into account the availability of the supplied units and its suitability with the predicting characteristics of the demand. This paper focuses on the segmented prediction within the framework of the Yield Management application. Biometrics as a Measure for Seamless
and Secure Checkin Procedures Today the most common trend in optimizing ground processes lies in airlines encouraging passengers to move towards more off-site CKI-channels. But while the operational and cost savings have certainly been significant, up to now the other half involving on-site passenger verification at Security and the Gate has not fully been covered. This absence of a complete end-to-end solution remains the critical gap to achieving efficient handling and movements of passengers. This presentation will address how Biometrics can serve to fill this gap through presenting characteristics of biometric systems, simplifying passenger handling and opportunities and threats of implementing biometric technologies. The Route Less Traveled Intra-regional air transport developments, in the light of planned EU enlargement in the Balkans, call for exploring the most suitable economic models as well as a strategy for regional aviation integration. The result could be the first regional ‘no-frills’ carrier in the Balkans based on the Southwest model and its EU derivatives. By exploiting benefits from the air transport liberalisation in the SEE (South-East Europe) region and through strategic analysis of current airlines’ route network and demand, new OD markets supporting a low fare airline model will arise. Interval of useful ranges for aircraft11 will be used for inter-modal accessibility range comparison in order to justify new routes to be flown by this low-fare carrier, as a result of modal shift from road to air transport. Also, secondary airports such as Batajnica in Belgrade, or additional terminals (Zagreb) will be examined to determine the most strategically appropriate hub in the SEE region for this regional ‘no-frills’ airline. The ultimate goal of this paper is to explore major conditions needed for the establishment of such carrier and to examine its financial, operational and economic viability.
Keynote Speakers (invited)
Call for Technical PapersThe 48th 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 21-26 September 2008 in Montreal, Quebec. 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 30-45 minutes, inclusive of minutes for Q&A. Symposium proceedings will be published by January 2009. *Your registration discount is based
on the conditional acceptance of your proposed presentation for the
symposium technical program, and only applicable to the primary
presenter. Deadlines and Requirements Abstract Submissions - 15 August 2008 (extended) Please send all submissions via the online form. 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. Paper/Presentation Submissions - 07 September 2008 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 2008:
Inquiries about Technical Program Tim
Jacobs |
|||||||||
Welcome | Agenda | Registration | General Information | Hotel Information | Host City |
||||||||||