Scheduling And Revenue Management Process


Published on

Scheduling And Revenue Management Process

Published in: Education, Business, Technology
  • Be the first to comment

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • Scheduling And Revenue Management Process

    1. 1. Scheduling and Revenue Management Process Integration: Benefits and Hurdles Timothy L. Jacobs, Elizabeth Hunt and Matt Korol Operations Research and Decision Support American Airlines May 2001
    2. 2. Presentation Overview <ul><li>Process Integration - What theory tells us. </li></ul><ul><li>Practical First Steps and Their Impact </li></ul><ul><ul><li>Consistent Scheduling and Revenue Management (O&D FAM). </li></ul></ul><ul><ul><li>O&D-based Demand Driven Dispatch (D 3 ) </li></ul></ul><ul><li>Benefits and Hurdles to Implementation </li></ul><ul><li>Summary </li></ul>
    3. 3. Airline Business Overview Scheduling Product Pricing Yield Management Sales and Distribution Long- Term Short- Term Strategic Tactical Customers
    4. 4. Typical Scheduling and RM Process Flight Scheduling (Leg-based) Revenue Management (O&D-based) Time 12 + Months 9-6 Months DOD 3 Months 45 Days O&D Demand Forecasts Revenue Management Process & Controls Capacities / O&D Forecasts Network Scheduling & Planning Flight Demand Forecasts Fleeted Schedule (Fixed Capacities) Data Source Data Source Informal Feedback
    5. 5. Proposed Integrated Process Causal Effect Data Date Specific Data Steady-State Industry Forecast O&D Daily Forecast O&D Time Series Forecast Revenue Management Process & Controls O&D Network Planning O&D-based Scheduling Near-term Aircraft Assignment (D 3 Process) Flight Scheduling Revenue Management Forecasting Data Sources Time 12 + Months 9-6 Months DOD 3 Months 45 Days Forecasts Forecasts Controls/Capacities O&D Forecasts/Capacities Forecast Data and Control Information
    6. 6. <ul><li>Provides a better balance between supply and demand and improves current practice by explicitly considering passenger flows in the scheduling process. </li></ul><ul><ul><li>Multiple O&Ds </li></ul></ul><ul><ul><li>Multiple Classes </li></ul></ul><ul><li>Consistent with Yield Management seat allocation and controls </li></ul><ul><li>Extensible to consider network recapture and pricing effects </li></ul>Consistent Scheduling and RM Benefits - O&D Fleet Assignment
    7. 7. Consistent Scheduling and RM Benefits - Theory Integrated Scheduling & Revenue Management Process No Revenue Management Revenue Management Only Reference: Jacobs, Ratliff and Smith;1997, 2000
    8. 8. Extension to Consider Pricing Effects No RM O&D RM O&D RM & Pricing O&D Fleeting and RM O&D Fleeting, RM & Pricing Reference: Jacobs, Ratliff and Smith;1997, 2000
    9. 9. <ul><li>Estimate O&D market forecasts. </li></ul><ul><li>Fleet schedule with a Segment-based Fleet Assignment Model (Leg-FAM). </li></ul><ul><li>Improve fleeted schedule using O&D FAM application. </li></ul><ul><li>Evaluate Leg-FAM and O&D FAM schedules using the O&D revenue mix model. </li></ul>O&D Fleeting and RM Benchmark Process - Practice O&D Forecast Leg FAM O&D FAM O&D Evaluation: Revenue Mix
    10. 10. O&D Fleeting and RM Benchmark <ul><li>General Information </li></ul><ul><ul><li>4,500 flight legs. </li></ul></ul><ul><ul><li>26 sub-fleets. </li></ul></ul><ul><ul><li>800 aircraft. </li></ul></ul><ul><ul><li>150,000 total O&D markets (Including International Markets). </li></ul></ul><ul><ul><li>No Jet-Prop Swaps. </li></ul></ul><ul><ul><li>International Fleeting Maintained. </li></ul></ul>
    11. 11. O&D Fleeting and RM Benchmark Results
    12. 12. Observations and Conclusions <ul><li>Benchmark results using existing forecasting methods and a consistent O&D Fleeting and RM approach illustrate significant potential benefits over segment-based FAM. </li></ul><ul><li>Additional benchmarks showed annual improvements ranging from 0.54% to 0.77% of revenue. </li></ul><ul><li>O&D Fleeting and RM process provides a better balance between available resources (capacity) and the O&D-based demands. </li></ul><ul><li>O&D Fleeting produces a schedule fleeting consistent with the RM process used to manage the seat inventory. This provides better opportunities to increase the overall schedule yield. </li></ul><ul><li>Potential benefits from a consistent O&D Fleeting and RM process will increase as forecasting capabilities improve. </li></ul>
    13. 13. <ul><li>Objective: Increase overall profitability by making strategic near-term aircraft swaps between crew compatible equipment. </li></ul><ul><li>Driving Forces: </li></ul><ul><ul><li>Paradigm shift: Many airlines fleet the schedule using leg-based methods while managing the seat inventory using O&D-based methods. This leads to an inconsistent matching of supply and demand. </li></ul></ul><ul><ul><li>Daily forecast variability: D 3 exploits opportunities created by the systemic daily variation of ODF demand flowing through the network. These effects are not captured when schedules are built using typical day forecasts. </li></ul></ul><ul><ul><li>Forecast Error: D 3 improves schedule profitability by using improved forecast data nearer the day of departure. </li></ul></ul>O&D-based Demand Driven Dispatch (D 3 )
    14. 14. <ul><li>Obtain remaining O&D Fare Class (ODF) demand forecasts, firm reservation holds, capacities and itinerary fares from RM for a specific reading day and departure date. </li></ul><ul><li>Improve fleeted schedule using O&D FAM and allowing only crew compatible RJ swaps. </li></ul><ul><li>Evaluate resulting schedule using the RM model and forecast data. </li></ul>Demand Driven Dispatch (D 3 ) Process RM Model O&D FAM Evaluation: Revenue Mix
    15. 15. Demand Driven Dispatch Benchmark <ul><li>Benchmark Information </li></ul><ul><ul><li>Reading Day 13. </li></ul></ul><ul><ul><li>Potential swaps: 566 candidate flight legs. </li></ul></ul><ul><ul><li>4800 total flight legs in schedule. </li></ul></ul><ul><ul><li>115,000 total O&D fare classes (Including International Markets) considered in analysis. </li></ul></ul><ul><ul><li>All other fleets held constant. </li></ul></ul>
    16. 16. D 3 Benchmark Results - Max Profit Measure* Input Schedule D 3 Solution Incremental Profit Gain (% of Revenue) 0.64 Switched Flights Segments Flown 114 RJ3 RJ4 230 * All measures are for a daily schedule 10:31 9:37 Utilization RJ3 RJ4 336 198 368 10:02 10:14
    17. 17. D 3 Parametric Analysis Results - Swap Limit Swap Limit Daily Profit Increase (% of Revenue) Cumulative Percent of Total 25 50 75 100 114 0.25 0.35 0.50 0.60 0.64 39% 56% 78% 94% 100%
    18. 18. A Closer Look - 25 Swap Limit Flight No. 1 2 3 Profit Change (% of Rev) 0.01 0.02 0.03 Reservation Holds 7 32 33 Incremental Traffic Input Output 7 4 3 8 9 10 Fleet Input Output RJ4 RJ3 RJ3 RJ3 RJ4 RJ4 Total Traffic Input Output 14 36 36 15 41 43
    19. 19. D 3 Benefits and Timing - What the theory tells us.
    20. 20. D 3 Benefits and Timing - The Practice
    21. 21. <ul><li>Results clearly illustrate the potential benefit associated with D 3 swaps of crew compatible aircraft near the day of departure. </li></ul><ul><li>D 3 effectively exploits the daily variations in ODF demand forecasts to identify revenue opportunities not realized during the schedule planning process. </li></ul><ul><li>D 3 provides an added degree of freedom to the RM process. This added flexibility allows the airline to adapt to better forecasts near the day of departure. </li></ul><ul><li>A portion of these benefits are likely due to inconsistencies between the scheduling and RM processes (Leg-based planning vs. O&D-based control). </li></ul><ul><li>Must account for M&E, crew and operational issues. </li></ul>D 3 Summary
    22. 22. Benefits and Hurdles to Integration <ul><li>Benefits: </li></ul><ul><ul><li>Consistent scheduling and RM processes can uncover significant revenue opportunities not realized in today’s process. </li></ul></ul><ul><ul><li>Implementation facilitates a natural and systematic feedback mechanism between scheduling and RM processes. </li></ul></ul><ul><ul><li>Provides opportunities for further process integration (pricing, M&E, Crew). </li></ul></ul><ul><li>Hurdles: </li></ul><ul><ul><li>Paradigm shift will require analysts to think about the scheduling problem in a much different way. </li></ul></ul><ul><ul><li>Process integration raises a host of process and schedule ownership issues that must be resolved. </li></ul></ul><ul><ul><li>Integration puts added emphasis on the importance of forecasting at the Leg and O&D level. </li></ul></ul><ul><ul><li>Timing of D 3 highly dependent on ability to market added capacity. </li></ul></ul>