PROUDCTION AND OPERATION MANAGEMENT.pptx

- Assessed research papers to evaluate the most apt model used by FedEX for optimum Airline cargo dynamic logistic algorithms - Operations research LPP with aircraft regression equations and incidence matrix for obtaining the most optimum cargo routes

1. PROUDCTION AND OPERATION MANAGEMENT THE HUB NETWORK DESIGN WITH STOPOVERS AND FEEDERS CASE OF FEDERAL EXPRESS BY -SOUVIK DAS (231250142) -TANEESHQ YOGI (231250150) -TARUN RAJ BODDU (231250152) -SUNANDAN KULDEEP (231250145) -SAURABH KUMAR (231250120) -SHUBHAM MALVIYA (231250131)

2. 01 02 03 04 05 1.INTRODUCTION Overview 2. COMPANY AND INDUSTRY BACKGROUND 3. MODEL DEVELOPMENT 4. TEST PROBLEM DATA 5. ANALYSIS

3. Hub Network Design: Real vs. Ideal Traditional (Ideal) Model: Pure hub-and-spoke network - every location connects directly to the hub. Real-World Networks: More complex with stopovers and feeders. Stopovers : Flights carrying multiple packages stop at multiple locations before reaching the hub (e.g., Las Vegas-Albuquerque-Memphis). Feeders : Smaller planes transfer packages to larger planes at intermediate cities (e.g., Vancouver-Seattle). Benefits of Stopovers & Feeders: Lower costs: Fewer flights with higher capacity = savings on planes, fuel, and crew. Efficient connections for low-volume locations: Smaller planes on short distances, larger planes on long distances. Introduction

4. Founded in 1973, FedEx is the leader in fast, reliable delivery for small, valuable packages. Industry growth driven by: Increased economic interdependence (need for fast goods movement) Technological advancements (efficient tracking and management) Deregulation (enabled hub-and-spoke networks) Hub-and-spoke network: Efficient sorting at Memphis hub, not individual flights. Minimizes delays by prioritizing speed over direct routes. Operates at night in smaller airports to avoid congestion. COMPANY AND INDUSTRY BACKGROUND

5. Operating Procedure: Late afternoon/evening pickup. Unsorted loading for 1-3 stop journeys. Sorting by destination at Memphis hub (10:30pm-1:30am). Early morning departure (3:00am-5:00am) for final destinations. Priority delivery by 10:30am via van fleet. Operating Procedure

6. Context and Research Gap Focus of Existing Research: Finding optimal locations for sorting hubs in air cargo networks (strategic decision). The Missing Piece: How to configure the network (flights, routes) for best efficiency after a hub location is chosen (operational decision). This Study: Bridges the gap by proposing a new model for designing efficient hub networks. Research Gap

7. Goal: Optimize air cargo collection (overnight) using a mix of direct flights, stopovers, and feeder routes for a pre-determined hub network. Key Assumptions: Fixed hub, airports, single hub structure, static routes/schedules, predictable cargo volume, and simplified flight costs. (Details in later discussion) Addressing Limitations: Hub location and served airports are strategic choices, while network design is tactical. A buffer is included for cargo volume fluctuations. Model Development direct connections pure hub-and-spokes hub-and-spokes with stopovers and feeders.

8. Model Development pure hub-and-spokes hub-and-spokes with stopovers and feeders.

9. Model Development pure hub-and-spokes hub-and-spokes with stopovers and feeders.

10. Model Development pure hub-and-spokes hub-and-spokes with stopovers and feeders.

11. Model Development pure hub-and-spokes hub-and-spokes with stopovers and feeders.

12. Network Components (1 Slide) Key Network Elements: Arcs : Direct connections between airports (e.g., A-C with aircraft type 1). Paths : Sequences of arcs to deliver packages to the hub (e.g., A-C-M with stopover). Routes : Sequences of arcs flown by the same aircraft (e.g., C-M via D). Relationships : A route can be used by multiple paths (e.g., route C-M-Z used by paths A-C-M and B-C-M). Some elements (e.g., C-M) can be arc, path, and route. All elements are directional (one-way movement). Notation : Indices for cities, paths, arcs, and routes. Incidence Matrices: Define relationships between elements (e.g., Path-Arc matrix shows which paths use an arc). Network Components and Notation

13. Network Components and Notation

14. TEST PROBLEM DATA Model Considers : Operational costs (distance, aircraft type) Daily fixed costs (labor per day of operation) Aircraft depreciation costs (based on route flying time) Investment: Model indirectly reflects investment through minimizing overall expense. Future model could incorporate fleet limitations. Constraints: Supply Constraint: All cargo from origin cities must be collected. Arc Flow Constraint: Limits cargo volume on a route to aircraft capacity minus a reserve margin (accommodates demand fluctuations). Binary Decision Variable: Activates/deactivates arc capacity based on cost and availability.

15. TEST PROBLEM DATA Craft type and OLS regression equations were estimated in the following form: as a function of the size of the intermediate city.’   T = a + b D +e where T is flight travel time in minutes; D is distance in miles; b is a regression coefficient representing the inverse speed in minutes per mile; a is the regression line intercept representing the estimated extra time it takes the aircraft to ascend to cruising altitude, attain cruising speed, and land; e is an error term.   The regression equations for the three aircraft types are shown below: DC lo-10 T = 28.96 + .117D (R’ = 0.979) B 727-200 T = 20.83 + .121D (R* = 0.971) Dassault Falcon T = 18.40 + .131D (R* = 0.984) DC lo-10 T = 28.96 + .117D (R’ = 0.979) B 727-200 T = 20.83 + .121D (R* = 0.971) Dassault Falcon T = 18.40 + .131D (R* = 0.984)

16. A Rule-Based Process Focus : Generating candidate flight paths for cargo collection. Key Rules: Maximum Path Time : Paths limited to a duration observed in real-world networks (e.g., 6.8 hours). Limited Stopovers: Max of 2 intermediate stops per path for efficiency. Stopover City Restrictions: Prioritizes cities within a specific angular range towards the hub. Considers large cities (1.5M+ population) outside the range for potential consolidation. Aircraft Size and Cargo: Model considers different aircraft sizes to match cargo volume on each route. Allows for allocation of packages from a single city across multiple routes for optimal efficiency.

17. Considerations Key Considerations in the Model: Aircraft and Arc Capacities Three aircraft types with varying payload capacities (3.3 - 59.9 tons). Arc capacity limited by assigned aircraft's payload. 10% reserve margin added for cargo volume fluctuations. Line-Haul Costs: Cost per mile for each aircraft type (from Air Transport World). Ignores volume flown (justified by average payload costs). Fixed Route Costs: Crew expenses and depreciation based on aircraft type and route distance. Crew cannot switch aircraft types due to licensing. Package Supply Volume Estimation: (Data limitations addressed) "Relative Supply Index" (RSI) considers business activity and population. Highest RSI route assumed to have 90% capacity utilization for each aircraft type. RSI used with conversion factors to estimate daily package volumes (tons).

18. Analysis Base Case vs. Alternatives This analysis compares the performance of three air cargo network designs: Base Case: This network, generated by the model, allows for stopovers and feeder routes (most efficient). Pure Hub-and-Spokes: This theoretical model only uses non-stop flights (less efficient). 73.59% more expensive to operate. 89.9% more lift capacity needed (low average payload - 43.73%).

19. Analysis Base Case vs. Alternatives This analysis compares the performance of three air cargo network designs: Base Case: This network, generated by the model, allows for stopovers and feeder routes (most efficient). Pure Hub-and-Spokes: This theoretical model only uses non-stop flights (less efficient). 73.59% more expensive to operate. 89.9% more lift capacity needed (low average payload - 43.73%).

20. Further Scope Network planning is a complex process best approached in stages. Sequential decision method: Stage 1 : Ground vs. Air Service: Decide for each city whether ground transportation or air service is more efficient based on factors like volume and cost. This defines "ground collection areas" and designates a central airport for each. Stage 2: Hub Location: Locate the optimal hub(s) to serve the cities identified in Stage 1. This could involve a model considering distance and minimum volume thresholds. Stage 3: Route Design: Design the most cost-effective network of feeder routes (connecting airports to hubs) and stopover routes (connecting flights with intermediate stops).

21. Initial Optimal

22. Thank You !