|
Presentations ( Based on submitted abstracts to date* )Advantages of Real-Time Revenue Management Real-time or Event-driven revenue management is a process by which airline seat controls are recalculated and published to control seat availability as passenger bookings are made through the GDS. The approach optimizes controls for any subset of flight legs in the network or the entire network simultaneously as bookings are made for itineraries offered on the airline network. This approach provides a great advantage over the current revenue management strategy in which inventory controls are at most re-optimized once every 24 hours via a batch computer job. By re-optimizing inventory controls more frequently, the revenue management process can compensate for the variability associated with passenger bookings and either restrict or make available seat inventory when appropriate. Air Cargo Overbooking Using a Discrete Show-up Rate
Distribution A key element to appropriate overbooking is an accurate estimation of the show-up rate of the bookings on hand. The main differences between the passenger and the cargo business, multi-dimensionality (weight and volume), smallest booking period, show-up rate higher than 100% (due to cargo over-tendering), higher data variability (from 1 kg. to 100 ton), suggest different approaches for estimating the cargo show-up rate. We show that the normal distribution, widely used for estimating the passenger show-up rate, does not always provide a good fit. We use a new method to fit a discrete distribution to the cargo show-up rate and conduct a comparison between the normal and the discrete estimates. We analyze the behavior of the overbooking levels when using the two different estimates. Airport Capacity Forecast in a CDM Approach Adverse weather is a major source for disruptions in airline operations. Therefore, in cooperation with Air Traffic Control The Netherlands and Amsterdam Airport Schiphol, KLM Royal Dutch Airlines has constructed a model that provides a forecast for the airport capacity for the next 24 hours. The forecast is based on the latest weather forecast, runway preferences and availability. The results will be used in a pilot based on Collaborative Decision Making (CDM). A Bi-level Programming Model of Price Competition This study proposes a bi-level model between an airline and it's competitors to develop strategies for the decision-making problem of pricing. This problem considers the measurement of decision-making aspects, which involves the profits of airline, competitors' market share, and travel demand. In the airline TNA study case, one outcome shows that if the airline TNA as the upper level and the competitors as the lower level, the strategies of airline TNA will promote the price. Another outcome shows that if airline TNA as the upper level and competitors as the lower level, the strategies of airline TNA will decrease the price. Bridging Business and Technology The airlines need to improve processes in order to reduce their costs and differentiate themselves in the face of emerging competition. This presentation showcases the value of creating a sense and respond enterprise and how to build such a system using an approach known as Process and Activity Flow and Framework (PAFF) ©. PAFF© closes the chasm between business and technology, utilizing a set of tightly integrated tools in an integrated development environment. It links all the stakeholders in an iterative cycle, building an improved system which is linked to business requirements, yet mindful of existing design constraints and budgetary cycles. Competitive Fare Shopping for Yield Management A problem with today's airline revenue management systems is that they are "closed systems"; little or no information is included about the carrier's current marketplace competitiveness in GDS and online website retail channels. Changes in OA schedules, fares, and availability make this marketplace competitive landscape highly dynamic. Although MIDT data provides some information along these lines, it is limited to bookings (not shopping activity). Most airlines recognize these limitations and are supplementing their revenue management systems with data from other sources (e.g. FareChase, SideStep and/or by checking online retail sites such as Expedia, Orbitz or Travelocity). The Sabre Labs and Research teams have developed an initial prototype of "Competitive Fare Shopping" using Travelocity fare search results. Fare shopping queries and associated results have been captured and aggregated into numerous different summary views. Such shopping activity at retail "storefronts" (either GDS's or websites) can help explain "why" a carrier's O&D fare class demand is moving in a particular direction. Such information is useful to both pricing and yield management analysts for better insights into the current competitive environment (by market and future date). Ultimately, it may also become an important data source for choice-based ODYM forecasting. Crew Pairings Recovery at Air France On the day of operations, when major disruptions occur (bad weather, airport close-down, etc.), one is left with very little time to modify the flight schedule and even less time to rebuild the crew pairings and rosters, thus leading to even more disruptions... For this purpose, Air France IT and OR departments have developped an application with which crew trackers can study potential or real disruption scenarios, rebuild the pairings and publish them to the crew rosters, all this within less than one hour. The presentation will describe the context of this project, the OR techniques which were developed for that purpose and the use of this application on real examples. Demand Functions and Pricing Models We investigate characteristics of demand functions for a closed system of mutually substitutable products, then formulate the demand functions which involve a linear complementarity problem (LCP). A pricing model based on the demand functioins will be proposed and analyzed. A possible simplification of the pricing model will be demonstrated. Depeaking Frankfurt: An Airline View Frankfurt Airport is highly slot-restricted. The arrival process is specifically critical due to the infrastructural nature of the airport. Congested airspace and infrastructure usually results in increasing blocktimes leading to a systematic distortion of the slot coordination process. The vicious circle is usually not seen in trajectory oriented airport simulation. A new method namely an iterative-stochastic simulation is introduced to project the delay impact of the schedule depeaking process resulting in the necessary proactive reduction of blocktime. Distribution of OD Revenue Management Controls Airlines use O&D level revenue management in an attempt to capture incremental revenue by gaining more control over inventory decisions. Unfortunately, much of the benefit gained by producing O&D controls is lost when the inventory is distributed via the traditional channels. Codeshares and alliances increase the risk of lost revenue opportunity. Attempts to stem this revenue leakage include journey data control, price as booked methodology, codeshare routers, and cascading. Evaluating Schedules with the Network Value Index A logit choice model yields a simple functional form for the computation of maximum expected utility. From this, a generalized "Network Value Index" for the comparison and evaluation of proposed enhancements to an existing airline network can be defined. This measure quantifies the value of the proposed enhancements from the passenger perspective, a necessary analysis to establish the economic viability of the changes. The underlying theory of the Network Value Index measure will be described. A case study will then be presented that shows the application of the concept to the evaluation of an actual long-haul schedule enhancement opportunity. Integrated Airline Planning The tactical planning process of an airline is typically decomposed into several stages among which fleeting, aircraft routing, and crew pairing form the core. In such a decomposed and sequential approach the output of fleeting forms the input to aircraft routing and crew pairing. In turn, the output to aircraft routing is part of the input to crew pairing. Due to the decomposition, the resulting solution is often suboptimal. We propose a model that integrates these three stages and thus yields a simultaneous solution to all three problems. We design two solution methodologies to solve the model. One is based on a combination of Lagrangian relaxation and column generation while the other one is the Benders decomposition approach. We give computational experiments for a variety of instances obtained by a major carrier. Investigation of Value at Risk in Airline Pricing Value-at-Risk (VaR) is a technique to estimate the likelihood that a given portfolio's losses will exceed a certain amount. It has been extensively used in financial risk-management since 90s to ensure the financial security and stability. At the meantime, by observing financial failures in airline industry, mainly due to high uncertainties(like terrorist attack and SARS outbreak), we believe it necessary to take risk-management into consideration in airline-pricing. In this presentation, we conduct the comparisons between the standard pricing approach of maximizing the total expected profit and the new Value-at-Risk approach under different market conditions, from which, we conclude that the new approach is better than the standard one in ensuring the security and stability in case of bad market condition. Moreover, we propose two methodologies for improving the stability in the industry: one is based on the standard pricing-approach and the other one on the VaR-approach. Monitoring and Controlling Key Performance Indicators Key Performance Indicators (KPIs) are used to measure the performance of an airline system, both from a financial and an operational point of view. Carmen Systems and MIT recently started a joint project with the objective to identify KPIs and understand the functional relationship between them and the important performance characteristic(s) of an airline. This research investigates the relation between existing financial and operational performance indicators, and if necessary, propose new KPIs for better understanding and controlling the parameters that influence the profitability of an airline. In this presentation, we will present some preliminary results for a selection of KPIs and show how these can be monitored and controlled. This will be done using a simulation model based on real operational data of an airline. Optimizing Airline Supply Chains The Airline industry faces unique challenges which presents tremendous opportunities for OR practitioners. The authors discuss an emerging area of optimization for airlines, namely Supply Chain Management. The complex global web of supply sources and customers presents a unique opportunity for airlines to optimize their supply chains, without the risk of compromising on service levels. The authors will also highlight the in-house skills and capabilities which Airlines possess to undertake such a complex and cross-functional engagement. A case study by Emirates Airline including potential benefits which are likely to accrue will be discussed. The authors expect to deliver insights into an emerging facet of optimization for Operations Research specialists to support their Airlines pursuit for sustained profitability. About Supply Chain Management: Supply chain management plays an important role covering the procurement, manufacturing and distribution cycle with each element touching multinational dimensions. This is primarily due to the fact that manufacturing plants, procurement divisions, warehouses and distribution centers are located miles apart thus opening up avenues for optimizing costs. Optimizing Codeshare Revenues Alliances and Codeshare agreements between Airline partners has proliferated in recent times. The growing number of Codeshare flights has raised a business need for the airline to be able to identify the “best” codeshare opportunities. Moreover, restrictions like Flight number limitations; regulations about the extent of Codesharing etc. further accentuate this need. We present an approach that identifies the best codeshare opportunities and revenue sharing scheme for a carrier. We extend the concept to establish a potential Nash Equilibrium between partners and seek a solution that optimizes revenues from an alliance perspective. Recurrent Training Scheduling The crew bidding and award process at Delta Air Lines resulted in a large number of trips falling into open time due to conflicts with vacation and training. The previous Pilot Working Agreement (PWA) didn't allow the flexibility to plan around these conflicts. The new PWA changed the training pay structure but also allowed the opportunity to schedule training after the bidding and award process was completed. To reduce the cost exposure under the new PWA, we developed a new model that not only minimizes conflicts with trips, but also maximizes simulator slot utilization and reserve crew utilization. In this presentation we describe the model and its benefits.Robust Fleet Assignment Fleet assignment models are used by many airlines to assign aircraft to flights in a schedule to maximize profit. The results of fleet assignment models affect subsequent planning, marketing and operational processes within the airline. Anticipating these processes and developing solutions favorable to them can further increase the benefits of fleet assignment models. We developed a fleet assignment approach that increases planning flexibility and reduces cost by imposing station purity, limiting the number of fleet types allowed to serve each airport in the schedule. Station purity makes fleeting solutions more robust relative to crew planning, maintenance planning and operations; however, station purity can significantly degrade computational efficiency. In order to more efficiently solve the station purity fleet assignment problem, we developed a column generation approach, Station Decomposition, which takes advantage of typical airline network structure. We further improved the performance of Station Decomposition by developing a primal-dual method that increases the solution quality and model efficiency. Station Decomposition solutions can be highly fractional; we develop a “fix and price” heuristic to efficiently find integer solutions to the fleet assignment problem. SIMAIR Simulation for Airline Operations SIMAIR (Simulation of Airline Operations) is a discrete event simulation of airline operations. It is a joint effort between Georgia Tech and National University of Singapore. SIMAIR is meant to be a research tool to help in evaluating effectiveness of an airline schedule recovery policy, and the robustness of a given schedule in operations. SIMAIR can simulate the operations and flow movement of aircrafts, crews, passengers and cargoes, taking into account the congestion of runways and airport gates. Prototypes of MINI-SIM (a deterministic look-ahead simulation which evaluates effectiveness of various potential recovery actions) and interactive, human-in-the-loop recovery have also been developed. SimDelta The R&D group at Delta Technology is developing a simulation, called SimDelta, to evaluate flight schedules by modeling Delta Air Lines’ flight operations. The first phase of the model focuses on the aircraft and crew resources and incorporates all significant sources of disruptions in daily operations. SimDelta models the Operations Control Center to capture some of the basic real-time decision making processes. In this presentation, we describe the basic model framework, which is easily extensible, and lay out the future course of development. SkyBook the Less Paper Cockpit Solution Today, flight crews suffer from the density and complexity of constant information overflow. Appropriate information has to be filtered to take the right decisions. Lufthansa Systems' Skybook is a mobile platform solution based on any standard notebook/TabletPC. This allows reducing significantly the amount of paper in airplane cockpits and gives flexible access to all relevant documents and to up-to-date operational data. It allows to introduce lean processes thus saving money. Flight preparation tasks can be performed, documented online and stored on the SkyBook. It is the key for taking every aspect of flight operations to a higher level of efficiency. State of the Art No Show and Standby Models This study models airline passengers' no-show and standby behavior using outbound and inbound itinerary information. We describe factors that influence no-show and standby behavior in continental U.S. markets using multinomial logit and nested logit models. Our model accounts for early standby opportunities on other itineraries flown by the carrier of interest, i.e., it incorporates information about booking levels on the carrier's itineraries departing on the same day. Empirical and validation analyses highlight the importance of distinguishing between outbound and inbound itineraries. We conclude by describing a framework carriers can use to implement our no-show model. Task Based Optimal Roster Generation Rostering and Manpower planning for airline ground operations has typically been done on a "typical" week of flight schedules and repeated for as long as required. Current market dynamics, however have forced airlines to vary their schedules across weeks and even days without necessarily "repeating". Under such conditions, a workload plan based on a "typical" week could be highly inefficient or impossible to operate. Moreover, Roster generation is typically done using a phased approach of generating a shift cover, rosters and task assignment, using "optimal" solutions at each stage. We present an integrated approach to generate globally optimal rosters factoring in all necessary constraints, and demonstrate quantifiable benefits of this approach. The Aircraft Maintenance Recovery Problem Unexpected events alter the flight schedule during day-to-day airline operations. Aircraft controllers react to these events by canceling or delaying flights and swapping aircrafts with the aim of recovering the flight schedule as quickly as possible. The real-time alteration of aircraft routes many times leads to aircraft missing their scheduled maintenance stops in the near future. After the schedule has been recovered and the airline is back to normal operations, the aircraft routers need to ensure that every aircraft honors the scheduled maintenance events at one of the maintenance stations. In this presentation we will describe the problem, propose a solution approach and present numerical results based on data from a major US airline. The Case for Robust Schedules Airlines must delay or cancel flight legs when airport capacity is reduced or when aircraft and/or crew are not available. The resulting disruptions to aircraft routings, crew pairings and passenger itineraries often propagate through the network and can last for many days. The severity and speed with which disruptions propagate is a function of the type of schedule that is developed. A schedule with very tight aircraft, crew and passenger connection will by its very nature be brittle while a schedule with ample slack will absorb delays. Airlines have sought to reduce the cost of operation by developing schedules with little or no slack because there are significant cost implications to having slack in the schedule. However, the cost of operation in an uncertain operating environment where flight times and airport capacity can vary significantly might actually increase as the planned cost decreases. In this presentation I will make the case for the development of more robust schedules through a review of airline performance statistics and present results of recent research that illustrates the benefits with little or no additional planned cost of “having a little fat in the schedule” provided it can be allocated only to those connections that need it most.
|
|||||||||||
|
Welcome | Agenda | Registration | General Information | Hotel Information | Host City |