Presentation is loading. Please wait.

Presentation is loading. Please wait.

PODS Update Large Network O-D Control Results Peter Belobaba and Seonah Lee Massachusetts Institute of Technology AGIFORS RM STUDY GROUP MEETING New York.

Similar presentations


Presentation on theme: "PODS Update Large Network O-D Control Results Peter Belobaba and Seonah Lee Massachusetts Institute of Technology AGIFORS RM STUDY GROUP MEETING New York."— Presentation transcript:

1 PODS Update Large Network O-D Control Results Peter Belobaba and Seonah Lee Massachusetts Institute of Technology AGIFORS RM STUDY GROUP MEETING New York City March 22-24, 2000

2 2 Outline Description of New Large PODS Network Standardization of RM and O-D Method Parameters  DAVN parameters – re-optimization, virtual bucket definition  Re-optimizing rate for bid price methods (HBP and PROBP) Results: O-D Revenue Gain Comparisons  Impacts of Average Load Factors and Distributions Overview of Additional PODS Studies  Use of Path-Based (ODF) Forecasts in Leg/Bucket RM  Introduction of Cancellation and No-Show Behaviors  Recovery of RM Methods from Sudden Demand Shocks

3 3 Characteristics of Large Network 40 spoke cities with 2 hubs, one for each airline 20 spoke cities on each side, located by geographical coordinates of actual US cities Distance to 1514 miles to the hub from spoke cities Unidirectional -- West to east flow of traffic Inter-hub services -- one for each direction, for each bank, for each airline 3 banks starting at 10:30, 14:00, 17:30 for each airline hub 252 flight legs, 482 O-D markets, 4 fare types per market

4 4 Geographical Layout H1(41) H2(42)

5 5 Standardization of O-D RM Methods “Generic” RM method parameters defined 3 years ago for smaller PODS networks (6-10 cities):  4 fare classes for Base Case EMSRb Control  6 virtual buckets per leg for GVN, HBP and DAVN  Network-wide virtual range definitions  Varying re-optimization rates for bid price methods For new 40-city network, we updated RM methods:  “Standard” definitions to better reflect actual and feasible implementations of each method

6 6 Standardized RM Method Parameters FCYM -- Fare Class Yield Management  4 fare classes grouped by yields and fare restrictions  Leg/class demand data and forecasting  EMSRb limits -- Re-optimize at 16 checkpoints GVN -- Greedy Virtual Nesting  ODFs mapped to 8 virtual buckets based on total itinerary fare values  Network-wide virtual ranges for all legs  Leg/bucket demand data and forecasting  EMSRb limits -- Re-optimize at 16 checkpoints

7 7 Standardized RM Method Parameters HBP -- Heuristic Bid Price  Like GVN, ODFs mapped to 8 virtual buckets based on total itinerary fare values  Same network-wide virtual ranges for all legs  Leg/bucket demand data and forecasting  EMSRb booking limit control for local (one-leg) itineraries -- re-optimized 16 times before departure  “Bid price” control for connecting requests based on current EMSR values of last seat on each leg: Re-optimized daily over 63-day PODS booking period

8 8 Standardized RM Method Parameters DAVN -- Displacement Adjusted Virtual Nesting  ODFs mapped to 8 virtual buckets based on displacement adjusted “network” revenue values: Network Value = ODF Fare - Displacement Cost  Leg Displacement Costs estimated by shadow prices of deterministic network LP optimization Network re-optimized at each checkpoint (16 times) Leg-specific virtual bucket range definitions  ODF demand forecasting (rolled up to leg/bucket)  EMSRb control of leg/buckets checkpoints

9 9 Standardized RM Method Parameters PROBP--Probabilistic Network Bid Price  Nested probabilistic network convergence algorithm developed at MIT (Bratu, 1998)  Involves “prorating” total ODF value to legs traversed: Requires ODF data demand forecasts Estimates “critical EMSR operator” for each leg by accounting for complete nesting of ODF availabilities  Critical EMSR values used as additive bid prices for local and connecting path requests Re-optimized daily over 63-day PODS booking period

10 10 Summary of New RM Parameters Base Case Fare Class YM effectively unchanged Enhancements to virtual bucket methods:  Number of virtual buckets increased to 8  More frequent network displacement optimization and leg- specific virtual re-bucketing for DAVN  Represents “advanced” implementations of DAVN More realistic bid price re-optimization frequency:  Airline consensus that daily bid price updates are feasible in larger networks  Theoretically more frequent updates might be misleading

11 11 Demand and Load Factors Simulated Under FCYM Base Case, simulated demand factors led to network ALFs from 70% to 87%  Load factor distributions compared well with system data provided by 2 airlines Local traffic represents 37 to 40% of total load by flight leg, on average:  Varies by demand factor and RM methods used Differences in load factors by connecting bank at each hub:  Highest for mid-day bank, lowest early in morning

12 12

13 13 ALFs by Hub Connecting Bank 3 banks per day offered at each airline’s hub:  Range of ALFs and revenue gains for each RM method  Most realistic traffic characterization in PODS to date

14 14 Revenue Gains over FCYM (Competitor uses FCYM)

15 15 Comparison of O-D Revenue Gains Relative performance in line with smaller network:  Small gains for GVN, negative at higher demands  HBP revenue improvements over “greediness” of GVN  DAVN and PROBP perform best, gains of 1% or more But, overall % gains of O-D methods are lower:  New network not designed to be “O-D friendly”  Each demand factor includes a range of ALFs by bank, with lower % gains for lower demand banks  More path choices without airline preferences or re-planning disutilities result in greater passenger shifts among paths

16 16 Revenue Gains by Connecting Bank (Network ALF=83%, Competitor uses FCYM)

17 17 Competitive Impacts of O-D Methods (Network ALF=83%, Competitor uses FCYM)

18 18 Competitive Impacts of O-D Methods O-D control can have substantial revenue impacts on competitor:  Continued use of FCYM against O-D methods results in revenue losses for Airline B  Interesting is GVN result, where Airline B’s revenue loss is greater than Airline A’s gain  Still not a zero-sum game, as revenue gains of Airline A exceed revenue losses of Airline B  Other simulation results show both airlines can benefit from using more sophisticated O-D control

19 19 Lessons from Larger Network Demand characteristics affect O-D benefits:  No explicit effort to design “bottleneck” legs that favor GVN  More realistic distribution of load factors across legs  Different load factors for connecting banks by time of day  Misleading to focus comparisons on peak connecting banks Characterization of O-D methods also critical:  More sophisticated DAVN parameters, more realistic PROBP re-optimization frequency  Robustness of DAVN even with periodic re-optimization O-D control has important competitive impacts

20 20 Large Network in PODS: Next Steps Alternative demand and network characteristics:  Proportion of local vs. connecting O-D demand  Load factor distributions  Business vs. leisure traffic mix Impacts of passenger choice disutility parameters:  Increase re-planning costs for changing preferred times  Modify airline preference factors from 50/50  Introduce path quality options (non-stops) and disutilities Less structured and more “realistic” O-D fares:  Not necessarily tied to O-D market distances

21 21 Overview of Other PODS Studies Path-Based (ODF) Forecasting in Leg-Based RM Introduction of Cancellation and No-Show Rates Impacts of Sudden Demand Shocks Competitive Studies Planned and Under Way

22 22 Path-Based Forecasting in Leg RM Preliminary results show potential gains from use of path-based (ODF) forecasts in leg-based RM:  ODF database to keep historical booking data  Tested simple moving average “pick-up” forecasts with “booking curve” unconstraining  ODF forecasts “rolled up” to leg/class or leg/bucket ODF forecasts not necessarily more “accurate”:  Error relative to mean forecast is large due to small numbers  But ability to unconstrain demand by ODF path appears to contribute in large part to revenue gains

23 23 Example: Path Forecasts for Leg RM (Previous Large Network ALF=75%)

24 24 Cancellation and No-show Rates Over past several months, we have incorporated cancellation and no-show processes into PODS:  “Memory-less” daily cancellation probability  Gaussian distributions of no-show rates at departure  Probabilistic overbooking model to determine AUs Neither process has a large impact on revenue gains of O-D methods:  Relative performance of methods stays the same at similar load factors; O-D methods do slightly better at lower ALFs Now testing gross vs. net booking forecast models

25 25 Impacts of Sudden Demand Shock Simulated “overnight” demand shifts of +/- 20%:  Extreme test of robustness of each RM method to changes in actual demand vs. forecast  Compared percentage revenue gains of each method vs. FCYM before and after demand shock After 20% sudden demand decrease:  GVN benefited, showing immediate revenue increase  DAVN and PROP suffered, due to over-forecasts by ODF  HBP maintained relative revenue gains  Relative performance stabilized after samples

26 26

27 27 Competitive Studies with PODS Introduction of third “new entrant” airline in one or more spoke-hub local markets:  What are impacts on hub carrier that uses leg vs. O-D RM?  What are “rational” vs. “predatory” responses by hub carrier in terms of prices, capacity and RM controls? System-wide reduction of aircraft capacity (6%?) by one hub airline to increase legroom:  Revenue and load impacts with leg-based vs. O-D RM?  What increase in airline preference is needed to make up for revenue losses?

28 28 Summary: PODS RM Research After four years of development, PODS network is now approaching “realistic” characterization. Change in recent emphasis of PODS simulations:  Away from O-D method “competitions”  Towards understanding major impacts on RM performance Ability to simulate larger networks opens up even greater potential for PODS research:  Airline alliances and other competitive strategies  Impacts of pricing and schedule changes on RM methods  Inclusion of scheduling and fleet assignment models

29 29 PODS Revenue Management Research at MIT MIT PODS Consortium of 6 international airlines Major accomplishments in past year:  Expansion of PODS network cities, 2 airlines, multiple banks per day  Establishment of “implementable” O-D methods  Focus on sell-up models and interaction with forecasts  Impacts on RM method performance of forecasting, demand shocks, fare structures, cancellations  New competitive studies involving RM Alliance RM Strategies Impacts of New Entrant Airlines


Download ppt "PODS Update Large Network O-D Control Results Peter Belobaba and Seonah Lee Massachusetts Institute of Technology AGIFORS RM STUDY GROUP MEETING New York."

Similar presentations


Ads by Google