Introduction to Operations Management.

Transcription

1 Introduction to Operations Management 1

2 Introduction to Operations Management Introductions 2. Housekeeping a. SloanSpace b. Course Introduction c. Professional Standards 3. Concepts & Nokia 4. Course Outline 5. Next Time a. Sega b. CPM

3 Housekeeping for Operations Management 3 1. Course Materials: Course packet E.M. Goldratt and J. Cox, The Goal: A Process of Ongoing Improvement, North River Press, 2nd Rev. Ed., The Memory Jogger, Goal/QPC, Grading Class participation: 30% First case write-up 20% Second case write-up 25% Third case write-up 25% 3. Professional Standards Academic Integrity-- Do your own work Behavioral Integrity -- Do unto others...

4 Three Foundational Components of Operations Management 4 Product Development Process Design & Management Supply Chain

5 5 Product Development Product Design -Voice of the Customer What is the role of product design in the demand and supply issues faced by Nokia and Ericsson? -Product/System Architecture Were problem chips integral or modular? Product Development -Project management & Cost -Design for Manufacturing How important was Nokia quickly redesigned some of its chips so they could be produced elsewhere? Technology Strategy Did product technology play a role in the differential performance of N & E?

6 Process Design & Management Process Design: Options & Assessment -Queueing Analysis -Capacity Analysis How did Nokia assess capacity in the crunch? How did they change capacity? -Uncertainty Analysis How did each company prepare for difficult-to-anticipate events? Inventory Systems Did N&E operate Just-in-Time, or did they hold big stores of chips waiting just in case? Production Control Was Nokia s software the principal instrument of control? How did they monitor the situation? ERP/Software/Internet Was Nokia s software the principal instrument of communication? Operations Excellence - Continuous Improvement - Just-in-Time - Quality Management (SPC, 6σ)

7 7 Supply Chain Strategic Supply Chain Design -Make Vs. Buy Did sourcing strategy play a role in the differential performance of N & E? -Supplier Selection, Sourcing Single vs. Dual sourcing Supply Chain Management -End-to-end coordination Do we see here examples of integrated enterprise? -Supplier Relations hard-nosed, polite, hostile, collaborative? Delayed Differentiation

8 Companies and Industries we will cover 8 Product Electronics & SW Process Autos: Electronics: Financial Food Retailing Food Processing Air Transport Health Care: Software: Supply Chain Electronics: Fashion Apparel Food Distribution esupply Sega Toyota Dell, Cisco, Quanta Bank of America Burger King National Cranberry Alaska Air University Health Sega, SAP (Vandelay), Oracle (Cisco) Nokia, HP Sport Obermeyer Barilla Pasta Webvan

9 Course Outline 9 1 Introduction Course Introduction Trial by Fire, powerpoint on Ops Strat 2 Product Dev Dreamcast/Sega Chap 8 in Clkspd on 3-DCE, ABC's of CPM 3 Operations Burger King Types of Processes, EOQ, Newsvendor 4 Strategy Inventory Mgmt Inven probs, Relevant costs, Whirlwind/Web, Dell/Conqueror, Laptop King 5 Process Alaska Airlines Levitt 6 Technology Webvan They've got mail. 7 Cisco MRP note, ERP Technology Note 8 Process Process Flow Models Queueing Note & Inventory Buildup 9 Analysis National Cranberry 10 Univ Health Service 11 Process Quality Mgmt Deming, Juran, Crosby; 6sig, Berwick, Memory Jogger 12 Quality Toyota Lean Production, Karmarkar 13 The Goal 14 Bank of America Hammer & Cole Articles 15 Supply Hewlett-Packard SMR paper 16 Chain Barilla SPA 17 Sport Obermeyer 18 Wrap-Up Wrap-up

10 Clockspeed: The Dimension of Time on Operations Management Study the Industry Fruitflies Evolution in the natural world: FRUITFLIES evolve faster than MAMMALS evolve faster than REPTILES THE KEY TOOL: Cross-SPECIES Benchmarking Evolution in the industrial world: INFOTAINMENT is faster than MICROCHIPS is faster than AUTOS evolve faster than AIRCRAFT evolve faster than MINERAL EXTRACTION THE KEY TOOL: Cross-INDUSTRY Benchmarking of Dynamic Forces of Dynamic Forces 10

11 INDUSTRY CLOCKSPEED IS A COMPOSITE: OF PRODUCT, PROCESS, AND ORGANIZATIONAL CLOCKSPEEDS 11 Mobile Phone INDUSTRY CLOCKSPEED THE obile Phone product technology THE Mobile Phone PRODUCTION PROCESS process technology THE Mobile Phone MANUFACTURING COMPANY organization

12 Mobile Phone System CLOCKSPEED is a mix of Transmission Standards, Software and Handsets Mobile Phone System 12 TRANSMISSION STANDARD slow clockspeed OPERATING SYSTEM slow clockspeed SOFTWARE APPLICATIONS medium clockspeed SERVICES fast clockspeed HAND SET fast clockspeed ISSUE: THE FIRMS THAT ARE FORCED TO RUN AT THE FASTEST CLOCKSPEED ARE THE MOST LIKELY TO STAY AHEAD OF THE GAME.

13 Clockspeed drives Business Strategy Cadence Dynamics between New Projects and Core Capability Development: PROJECTS MUST MAKE MONEY AND BUILD CAPABILITIES CORE CAPABILITIES See Leonard-Barton, D. Wellsprings of Knowledge NEW PROJECTS (New products, new processes, new suppliers)

14 ALL COMPETITIVE ADVANTAGE IS TEMPORARY 14 Autos: Ford in 1920, GM in 1955, Toyota in 1990 Computing: IBM in 1970, DEC in 1980, Wintel in 1990 World Dominion: Greece in 500 BC, Rome in 100AD, G.B. in 1800 Sports: Bruins in 1971, Celtics in 1986, Yankees no end The faster the clockspeed, the shorter the reign

15 ARCHITECTURES IN 3-D INTEGRALITY VS. MODULARITY Integral product architectures feature close coupling among the elements - Elements perform many functions - Elements are in close spacial proximity - Elements are tightly synchronized - Ex: jet engine, airplane wing, microprocessor Modular product architectures feature separation among the elements - Elements are interchangeable - Elements are individually upgradeable - Element interfaces are standardized - System failures can be localized - Ex: stereo system, desktop PC, bicycle

16 SUPPLY CHAIN ARCHITECTURE Integral supply-chain architecture features close proximity among its elements - Proximity metrics: Geographic, Organizational Cultural, Electronic - Example: Toyota city - Example: Ma Bell (AT&T in New Jersey) - Example: IBM mainframes & Hudson River Valley Modular supply -chain architecture features multiple, interchangeable supplier and standard interfaces - Example: Garment industry - Example: PC industry - Example: General Motors global sourcing - Example: Telephones and telephone service

17 DESIGNING ARCHITECTURES FOR PRODUCTS & VALUE CHAINS: THE NEED FOR ALIGNMENT VALUE CHAIN ARCHITECTURE (Geog., Organ., Cultural, Elec.) PRODUCT ARCHITECTURE INTEGRAL MODULAR INTEGRAL Jet engines Microprocessors Mercedes vehicles Automotive Supplier Parks MODULAR Polaroid Nortel, Lucent Personal Computers Bicycles Chrysler Vehicles Cisco

18 Product Process Supply Chain 18 Design Detailed Perform. Specs &Funct. Architect. Modular vs. Integral Unit Processes Tech. & Equip. Mfg.Syst Functnl Cellular. S.C. Architect. Orgs Set & Alloc. of Tasks Logistics &Coord System Auton vs. Integrated - Focus -Architecture - Technology A 3-D CE decision model illustrating the imperative of concurrency

19 DESIGNING ARCHITECTURES FOR PRODUCTS & VALUE CHAINS: MODULARITY VS. OPENNESS ARCHITECTURAL PROPRIETARINESS CLOSED ARCHITECTURAL STRUCTURE INTEGRAL MODULAR Pentium Chip Mercedes Vehicles SAP ERP IBM Mainframes Microsoft Windows Chrysler Vehicles OPEN Linux Palm Pilot software & accessories Phones & service Web-based ERP INFORMATION ARCHITECTURE MUST REFLECT BUSINESS MODEL

20 All Conclusions are Temporary 20 Clockspeeds are increasing almost everywhere 3-D Concurrent Engineering must anticipate Industry and Value Chain Dynamics 3-D Concurrent Engineering is a key organizational competency Study of Fruit Flies can help with crafting strategy

21 Course Outline Introduction Course Introduction Trial by Fire, powerpoint on Ops Strat Product Dev Dreamcast/Sega Chap 8 in Clkspd on 3-DCE, ABC's of CPM Operations Burger King Types of Processes, EOQ, Newsvendor Strategy Inventory Mgmt Inven probs, Relevant costs, Whirlwind/Web, Dell/Conqueror, Laptop King Process Alaska Airlines Levitt Technology Webvan They've got mail. Cisco MRP note, ERP Technology Note Process Process Flow Models Queueing Note & Inventory Buildup Analysis National Cranberry Univ Health Service Process Quality Mgmt Deming, Juran, Crosby; 6sig, Berwick, Memory Jogger Quality Toyota Lean Production, Karmarkar The Goal Bank of America Supply Hewlett-Packard SMR paper Chain Wrap-Up Barilla SPA Sport Obermeyer Wrap-up Hammer & Cole Articles

22 15.760: Sega Dreamcast/CPM What are the key elements of the value chain for the home video game business? 2. What are the possible failure modes for Sega and its Dreamcast product? 3. What capabilities does Sega require to succeed? 4. What s the critical path for the CPM exercise? 5. Launch as scheduled or postpone? 6. What happened. Why? 7. Lessons learned.

23 15.760: CPM exercise Task Description Duration (weeksprecedence A Market Study 14 B External Developer Focus Group 10 C Feature Selection 4 A,B D Hardware Engineering 25 C E Operating System Devlpt. 16 D,B F Advertising Campaign 20 D,M,N G Supplier Selection & Negotiation 20 D H Component Inventory Buildup 45 G I Assembly Facility Setup 18 D J Finished Good Inventory Buildup 7 I,H K Library & Programmer Toolkit Devlpt. 12 E L External Development Support Setup 5 K M Internal Game Devlpt. 30 K N External Game Devlpt. 32 L,O O Platform Promotion 6 K P Publisher Selection & Negotiation 5 M Q Website Setup 20 M R Release Promotion Material Design 3 D,M,N S Distribution Channels Devlpt. & Negotiation 9 R T Carrier Selection & Negotiation 4 S U Launch Event Organization & PR 5 R V Hardware & Software Shipment 1 J,T,P,N W Launch! 0 V,U,Q,F 23

24 15.760: CPM exercise What is the critical path and total duration for this project? 2. Suppose that a marketing study finds that the Advertising Campaign can be shortened to 15 weeks without significant impact reduction. What is the new CP and total project duration? Suppose we would like to reduce the time until launch to 28 months, what would you recommend? 4. Bonus: Suppose that the duration of the development tasks E, M and N are now random variables following exponential distributions with means 16, 30 and 32 respectively. What is the probability that the total project duration will be more than 135 weeks?

25 Mkt Study A 14 B 10 Features Ext Dev Focus Grp C 4 Setup Factory D 25 E 16 I 18 O/S Dev Supp Sel & Neg G 20 HW Eng J 7 Comps H Inv 45 Int Dev SW Dev Kit M K O 6 Platform Mktg FG Inv Build Dev Supp L 5 N 32 Setup Website Q 20 P 5 Publisher Sel & Neg Ext Dev F 20 Advert Plan R 3 Release Promo Ship HW & SW U 5 V 1 Plan Launch Event Distr Neg 25 W 0 GO! S 9 T 4 Carrier Neg

26 0, 14 A 14 0,14 14, 18 C 4 0, 10 B 10 4,14 ES, EF LS, LF ,18 18, 43 D 25 18, 43 43, 61 I ,121 43, 59 E 16 43, 59 43, 63 G 20 56,76 59, 71 K 12 59, 71 71, 77 O 6 71, , 115 J 7 121,128 63, 108 H 45 76, , 101 M 30 79, , 76 L 5 72, , 121 Q , , 106 P 5 123, , 111 R 77, N 112, , , 129 F , , 125 V 1 128, , 116 U 5 124, , W 0 129, , 120 S 120, T 115, ,128

27 0, 14 A 14 0,14 14, 18 C 4 0, 10 B 10 4,14 ES, EF LS, LF ,18 18, 43 D 25 18, 43 43, 61 I ,121 43, 59 E 16 43, 59 43, 63 G 20 56,76 59, 71 K 12 59, 71 71, 77 O 6 71, , 115 J 7 121,128 63, 108 H 45 76, , 101 M 30 79, , 76 L 5 72, , 121 Q , , 106 P 5 123, , 111 R 77, N 112, , , 129 F , , 125 V 1 128, , 116 U 5 124, , W 0 129, , 120 S 120, T 115, ,128

28 Corporate Decision d elayed launch Supplier Performance NEC recovers NEC late NEC very late Market Dynamics launch Exhaus tively seek a production fix NEC recovers Demand H Demand L Demand H Demand L Demand H Demand L Effects on Relationships Retailers increase dedicated space Software development accelerate Win-Win-Win!!! Product is a dud Reinvigorate or Exit. Sales puts on the best face possible Pressure/help NEC Retailers unhappy with empty shelves Disgruntled Customers Software developers anxious Need to stimulate demand & NEC Product may be a dud Software developers may quit Production ramp-up crucial Must maintain good communications Slow start is overcome: win-win-win Recrimination mode Was product a dud or was lateness to market the killer? Deep trouble Try to mend relationships and make new plan or just exit. 28

29 INDUSTRY CLOCKSPEED IS A COMPOSITE: OF PRODUCT, PROCESS, AND ORGANIZATIONAL CLOCKSPEEDS 29 Mobile Phone INDUSTRY CLOCKSPEED THE obile Phone product technology THE Mobile Phone PRODUCTION PROCESS process technology THE Mobile Phone MANUFACTURING COMPANY organization

30 Mobile Phone System CLOCKSPEED is a mix of Transmission Standards, Software and Handsets Mobile Phone System 30 TRANSMISSION STANDARD slow clockspeed SOFTWARE APPLICATIONS medium clockspeed OPERATING SYSTEM SERVICES fast clockspeed HAND SET fast clockspeed slow clockspeed ISSUE: THE FIRMS THAT ARE FORCED TO RUN AT THE FASTEST CLOCKSPEED ARE THE MOST LIKELY TO STAY AHEAD OF THE GAME.

31 Product Process Supply Chain Design Detailed Perform. Specs &Funct. Architect. Modular vs. Integral Unit Processes Tech. & Equip. Mfg.Syst Functnl Cellular. S.C. Logistics Architect. &Coord Orgs Set System & Alloc. Auton vs. of Tasks Integrated - Focus -Architecture - Technology A 3-D CE decision model illustrating the imperative of concurrency

32 Course Outline, A Introduction Course Introduction Trial by Fire, powerpoint on Ops Strat Product Dev Dreamcast/Sega Chap 8 in Clkspd on 3-DCE, ABC's of CPM Operations Burger King Types of Processes, EOQ, Newsvendor Strategy Inventory Mgmt Inven probs, Relevant costs, Whirlwind/Web, Dell/Conqueror, Laptop King Process Alaska Airlines Levitt Technology Webvan They've got mail. Cisco MRP note, ERP Technology Note Process Process Flow Models Queueing Note & Inventory Buildup Analysis National Cranberry Univ Health Service Process Quality Mgmt Deming, Juran, Crosby; 6sig, Berwick, Memory Jogger Quality Toyota Lean Production, Karmarkar The Goal Bank of America Supply Hewlett-Packard SMR paper Chain Wrap-Up Barilla SPA Sport Obermeyer Wrap-up Hammer & Cole Articles

33 Three Foundational Components of Operations Management 33 Product Development (Sega, #2) Process Design & Management (Burger King, #3) Supply Chain (Nokia, #1, Dell, #4)

34 15.760: Burger King What are the operations objectives for Burger King? 2. Process flow diagram for hamburger sandwich production. 3. Where are the inventories? Why? 4. Peak hourly capacity vs. peak hourly demand for burger patties. 5. How does the management of operations relate to the company's method of competing in the marketplace? 6. What are the fastest clockspeed components of the Burger King value chain? 7. How well integrated are BK s product, process, and market?

35 BK: Process Flow Diagram for Sandwiches RAW BURGERS BROILER MATE BUNS & BURGERS STEAMER FIN- ISHED SAND- WICHES RAW BUNS BROILER CHEESE SANDWICH ASSEMBLY CONDI- MENTS TOMA- TOES MICRO- WAVE ORDER DELIVERY

36 BK: Peak Load Demand Capacity Sandwiches 4.3 weeks x 18% (Fridays) x 17.9% = month month (noon hour) 256 sandwiches, of which 48 are hamburgers peak hour 51 are cheeseburgers 24 are double cheeseburgers => peak demand = 147 burger patties/hour Each broiler chain cooks 8 patties => 480 patties minute hour (Assume other chain used for Whoppers) Case fact: max assembly rate = 200 burgers/hour 100 specialty sandwiches/hr Broiler utilization = 147/200 = 74% Bottleneck is assembly

37 Restaurant Operations Management What are the key DESIGN parameters for Burger King? A. Product B. Process Technology C. Facility D. Work System/HR System 2. What are the key PLANNING tasks for Burger King? A. Supply B. Demand C. Capacity/Workload 3. What are the key CONTROL processes for Burger King? A. Production Control B. Quality Control C. Process Control 4. What are the key IMPROVEMENT processes for BK? A. Quality Improvement B. Productivity Improvement C. Technological Improvement D. Systems Improvement

38 Some Characteristics of Services (vs. Manufacturing) 38 Intangibility - explicit and implicit intangibles We manufacture perfume; we sell hope. Perishability - an hour of non-production is an hour lost Airplane w/o spare part costs > $10K/hr Heterogeneity - inherent variability of service Each doctor s bedside care is unique Simultaneity - services are simultaneously produced and consumed A poor attitude by the server cannot be recalled

39 INDUSTRY CLOCKSPEED IS A COMPOSITE: OF PRODUCT, PROCESS, AND ORGANIZATIONAL CLOCKSPEEDS 39 Mobile Phone INDUSTRY CLOCKSPEED THE obile Phone product technology THE Mobile Phone PRODUCTION PROCESS process technology THE Mobile Phone MANUFACTURING COMPANY organization

40 Mobile Phone System CLOCKSPEED is a mix of Transmission Standards, Software and Handsets Mobile Phone System 40 TRANSMISSION STANDARD slow clockspeed SOFTWARE APPLICATIONS medium clockspeed OPERATING SYSTEM SERVICES fast clockspeed HAND SET fast clockspeed slow clockspeed ISSUE: THE FIRMS THAT ARE FORCED TO RUN AT THE FASTEST CLOCKSPEED ARE THE MOST LIKELY TO STAY AHEAD OF THE GAME.

41 Product Process Supply Chain Design Detailed Perform. Specs &Funct. Architect. Modular vs. Integral Unit Processes Tech. & Equip. Mfg.Syst Functnl Cellular. S.C. Logistics Architect. &Coord Orgs Set System & Alloc. Auton vs. of Tasks Integrated - Focus -Architecture - Technology A 3-D CE decision model illustrating the imperative of concurrency

42 15.760: Dell/Quanta/Inventory Admin: SloanSpace, Namecards, CourseIntro 2. Explain the logic of the EOQ model. How useful do you think it is? 3. Dellpaq problem 4. Explain the logic of the newsvendor model. How useful do you think it is? 5. Fruitfly TM problem 6. How would you characterize Dell s approach to inventory management? Are the math models relevant or informative?

43 Inventory Problems: Dellpaq Computers Relevant costs Fixed Costs: shipments $100,000 placing order 5hrs * $10/hr = $50 S = $100,050 Variable Costs: complete systems $3000 loading plus shipping $25 unloading employee.1 hr * $10/hr = $1 unloading equipment.1 hr * $50/hr = $5 arranging in storage.05 hr * $10/hr = $.50 C = $ Irrelevant costs Removing sets from storage, cost of processor and motherboard, surveillance/security equipment

44 Inventory Problems: Dellpaq Computers Annual demands: Carrying charge: R=300,000 K=20% EOQ = 2(300,000)(100,050) = ( )

45 Inventory Problems: Fruitfly TM Decision: D Demand: d~normal(150000,45000) Price: p=150 Cost: c=50

46 Inventory Problems: Fruitfly TM P(d Q) = r-c = = 2 r Z(.167) =.431 Q = 150, * 45,000 = 169,395

47 Dell Operations Management What are the key DESIGN parameters for Dell? A. Product B. Process Technology C. Facility D. Work System/HR System 2. What are the key PLANNING tasks for Dell? A. Supply B. Demand C. Capacity/Workload 3. What are the key CONTROL processes for Dell? A. Production Control B. Quality Control C. Process Control 4. What are the key IMPROVEMENT processes for Dell? A. Quality Improvement B. Productivity Improvement C. Technological Improvement D. Systems Improvement

48 The Strategic Impact of Development Partnership Design: Who let Intel Inside? 1980: IBM designed a new product and created a new set of development partners Customers IBM Intel Inside Intel Microsoft The Outcome: A phenomenally successful product design A disastrous market value impact (for IBM) 48

49 LESSONS FROM A FRUIT FLY: THE PERSONAL COMPUTER BEWARE OF INTEL INSIDE (Regardless of your industry) 2. MAKE/BUY IS NOT ABOUT WHETHER IT IS TWO CENTS CHEAPER OR TWO DAYS FASTER TO OUTSOURCE VERSUS INSOURCE. 3. DEVELOPMENT PARTNERSHIP DESIGN CAN DETERMINE THE FATE OF COMPANIES AND INDUSTRIES, AND OF PROFIT AND POWER 4. THE LOCUS OF VALUE CHAIN CONTROL CAN SHIFT IN UNPREDICTABLE WAYS

50 Vertical Industry Structure with Integral Product Architecture Computer Industry Structure, Microprocessors IBM DEC BUNCH Operating Systems Peripherals Applications Software Network Services Assembled Hardware All Products All Products All Products (Adapted from A. Grove, Intel; and Farrell, Hunter & Saloner, Stanford)

51 Horizontal Industry Structure with Modular Product Architecture Computer Industry Structure, Microprocessors Operating Systems Peripherals Applications Software Network Services Assembled Hardware Intel Moto AMD etc Microsoft Mac Unix HP Epson Seagate etc etc Microsoft Lotus Novell etc AOL/Netscape Microsoft EDS etc HP Compaq IBM Dell etc (Adapted from A. Grove, Intel; and Farrell, Hunter & Saloner, Stanford)

52 THE DYNAMICS OF PRODUCT AND VALUE CHAIN STRUCTURE: THE DOUBLE HELIX 52 See Fine & Whitney, Is the Make/Buy Decision Process a Core Competence?

53 Product Process Supply Chain Design Detailed Perform. Specs &Funct. Architect. Modular vs. Integral Unit Processes Tech. & Equip. Mfg.Syst Functnl Cellular. S.C. Logistics Architect. &Coord Orgs Set System & Alloc. Auton vs. of Tasks Integrated - Focus -Architecture - Technology A 3-D CE decision model illustrating the imperative of concurrency

54 15.760: Alaska Air Admin/Intro: PeaPod, Webvan 2. What are important characteristics of services vs. manufactured products? 3. In what sense must AA concurrently develop product & process? 4. What are the key operations tasks that AA must accomplish? 5. Can self-service through technology provide a sustainable competitive advantage for AA? 6. Why use hubs?

55 15.760: WEBVAN ASSIGNMENT Take the position of a consultant to Webvan. Write a memo to the CEO of Webvan, providing him with your assessment of the Webvan operations model and business model -- based solely on the information available in the case. Include in your assessment some analysis of the value proposition to customers as well as the value proposition to investors. (I.e., how profitable will Webvan likely be?) (Max length: 1500 words & six supporting exhibits.) We all know that Webvan went broke. Feel free to take the position either that the business could make money (and provide the assumptions and/or analysis and/or creative ideas that support your position) or that the business is hopeless. If you take the latter position, you still need to back up your assessment with an analysis of why you concluded what you did. (I.e., what specifically are the flaws in the business model or operations assumptions?)

56 Some Characteristics of Services (vs. Manufacturing) 56 Intangibility - explicit and implicit intangibles We manufacture perfume; we sell hope. Mgmt Issues: Expectations & Perceptions Perishability - an hour of non-production is an hour lost Airplane w/o spare part costs > $10K/hr Mgmt Issues: Preventation Processes & Mentality Heterogeneity - inherent variability of service & customer Each doctor s bedside care is unique, as is each patient Mgmt Issues: Hiring, Training, Listening, Customization Simultaneity - services simultaneously produced & consumed A poor attitude by the server cannot be recalled Mgmt Issues: Hiring, Training, Situation Control

57 Three Foundational Components of Operations Management 57 Process Design & Management (Burger King, #3) Product Development (Sega, #2) (Alaska Air #5) Supply Chain (Nokia, #1, Dell, #4)

58 Three Foundational Components of Operations Management Process Design & Management (Burger King, #3) Product Development (Sega, #2) (Alaska Air #5) customer coddling Dell Product Features Supply µp & modem speed CD ROM speed Chain MB (Nokia, DRAM & HD #1, Dell, #4) screen size order-to-deliv time features range fufillment accuracy 58 AA Product Features check-in time reservations help meals price flight frequency mileage awards route coverage baggage handling

59 Operations Tasks for Alaska Airlines DESIGN A. Product Features (See Above; Contrast with Southwest) B. Fleet Configuration C. Reservations System D. Human Resources System (Care of Employees; Esprit de Corp) 2. PLANNING A. System schedule; B. Hiring/Training of Pilots, Flight Attendants, Ground Crew C. Workforce scheduling, Aircraft scheduling & maintenance D. Pricing, Marketing, Promotions (Competitor Actions) 3. CONTROL A. Reservations, Check-in, & Flight processes B. Responses to unplanned events: weather, breakdowns, etc. C. Processing Customer Complaints 4. IMPROVEMENT A. Operations Costs; Capacity Utilization (Role of Technology) B. Customer Experience (Role of Technology) C. Reputation & Brand

60 Technology Strategy at Alaska Airlines Hypothesis: Self-service through technology provides sustainable competitive advantage. ***Eliminate, Simplify Automate*** 1. Culture of Innovation (among tech leadership) 2. IMAGE system customer service, employee satisfaction, flexibility with laptops, missed opportunity? 3. On-line booking, e-tickets, kiosks, redeem miles Reduce costs and Improve Customer Satisfac. 4. Navigation and Guidance Systems Is this profitable? Sustainable?

61 Why use hubs (e.g., Seattle)? Victoria Calgary Victoria Calgary 61 Seattle Seattle San Francisco Las Vegas San Francisco Las Vegas

62 15.760: Webvan Admin/Intro: PeaPod: tomorrow night 2. What are the important features of the product/service offered by Webvan? How does it differ from competing offerings?. 3. In what sense must Webvan concurrently develop product & process?. 4. What are the key operations tasks that Webvan must accomplish? 5. Can you make sense out of their financial model? 6. Why use hubs & spokes?

63 15.760: WEBVAN ASSIGNMENT Take the position of a consultant to Webvan. Write a memo to the CEO of Webvan, providing him with your assessment of the Webvan operations model and business model -- based solely on the information available in the case. Include in your assessment some analysis of the value proposition to customers as well as the value proposition to investors. (I.e., how profitable will Webvan likely be?) (Max length: 1500 words & six supporting exhibits.) We all know that Webvan went broke. Feel free to take the position either that the business could make money (and provide the assumptions and/or analysis and/or creative ideas that support your position) or that the business is hopeless. If you take the latter position, you still need to back up your assessment with an analysis of why you concluded what you did. (I.e., what specifically are the flaws in the business model or operations assumptions?)

64 Challenges of Service Interface: Grocery Stores vs. Webvan 64 Intangibility - customer expectations vs. perceptions Grocery Stores: quality, selection, ENVIRONMENT Webvan: quality, selection, DELIVERY Perishability - use it or lose it Grocery Stores: fresh foods (produce, meats, baked goods) Webvan: fresh foods & TRUCK CAPACITY Heterogeneity - inherent variability of service & customer Grocery: checkout people, counter people, customer needs Webvan: DELIVERY PERSON Simultaneity - services simultaneously produced & consumed Grocery: presentation in the store Webvan: DELIVERY TO THE HOME

65 Three Foundational Components of Operations Management 65 Process Design & Management (Burger King, #3) Product Development (Sega, #2) (Alaska Air #5, Webvan #6) Supply Chain (Nokia, #1, Dell, #4)

66 Three Foundational Components of Operations Management Process Design & Management (Burger King, #3) Product Development (Sega, #2) (Alaska Air #5, Webvan #6) Grocery Store Features selection price quality/freshness shopping environment Webvan Features selection price quality/freshness shop any hour never leave home choose delivery time save your time same day delivery fulfillment accuracy no lugging required Supply Chain (Nokia, #1, Dell, #4) 66 Who has the advantage on each dimension?

67 Hub Operations for Webvan ASSUMPTIONS & ANALYSIS, CASE EXHIBIT 2 A. 50,000 SKU s B. 2,920,000 orders/year C. $103.00/order D. ==> $300,000,000 revenue per hub per year E. Requires /50000= 58 big $100K = $5.8M 58x40x52= hours/year ==> /(365x24)= 14 trucks on road F. Requires 580 small trucks per $60K = $34.8M ==> 140 small trucks on the road G. If grocery business serves 20 stores per hub, they would need twice the big truck capacity, but no small trucks, ==> Webvan extra delivery cost ~ $34.8M-$5.8M=$29M ~ %9.67

68 Hub Operations for Webvan HUB PRODUCTIVITY A. 450 items/person/hr B. Assume 300,000,000 items/yr C. => 666,667 man hrs/yr = $10/hr D. => $666,667/spoke /yr in direct labor E. Assumed 8000 orders/day x 365 x $103 = ~ $300M 68 F. Summer 1999: Sales of $1.05M/month Avg order = $71 => 14,789 orders/month = 493 orders/day Management estimated that the hub was operating at less than 20% of its designed capacity. G. 26 warehouses & 260 spokes staffed by people at $40K/year => $1,538,461 in labor per spoke

69 Grocery Grocery Rev/store/yr 14,055,000 Webvan Webvan 69 Rev/spoke/yr 30,000,000 Inv turns Cost of goods ,372, ,140,000 Gross Margin ,682, ,860,000 Store related costs IT costs 0.02 HQ alloc 0.03 Warehouse alloc 0.07 warehouse labor warehouse opns total labor Home Delivery Total expenses Net Oper Profit >%2? >3%? %12??? 300, ,667 1,538,461 2,900,000

70 Operations Tasks for Webvan DESIGN 2. PLANNING 3. CONTROL 4. IMPROVEMENT

71 Matching Fulfillment Strategies with Product Features Demand uncertainty H L Computers Books & CDs L Furniture Grocery H Delivery cost Unit price Prof. David Simchi-Levi, MIT

72 15.760: Cisco/MRP/ERP Admin: Peapod, Questions/Feedback 2. What were the lessons of the Beer Game? 3. How does production control work in the beer game? 4. Explain the purpose and logic of MRP. 5. What is the function of Cisco s ERP system?

73 Volatility Amplification in the Supply Chain: The Bullwhip Effect Order Order Order Info Info Info Retailer Wholesaler Distributor Factory How does production control work in the Beer Game? Information lags Delivery lags SOLUTIONS: Over- and underordering Countercyclical Markets Misperceptions of feedback Countercyclical Technologies Collaborative channel mgmt. Lumpiness in ordering (Cincinnati Milacron & Boeing) Chain accumulations

74 Applying EOQ and Newsvendor models to set Reorder Points and Reorder Quantities (s,s) (ROP, ROQ), (min, max) 74 ROP SS Q Q Q Q Q = 2RS CK = ROQ (REORDER QUANTITY) ROP=Reorder Point = Expected Demand During the order lead time + safety stock = E{DDL} + SS Prob {DDL ROP} = Cu/(Co+Cu) Cu=Cost of Underage (r-c in newsvendor); Co=Cost of Overage (c in newsvendor) But, Co with nonperishables is c x cost of holding ROP=SS+E{DDL}; DDL = X1 + X XL; E{DDL} = E{L} x E{X} i.e., DDL has a mean of Expected lead time x Expected avg demand/unit time Variance{DDL}~Var{X} x E{L} + Var{L} x E{X 2 }

75 What is the Purpose and Logic of MRP? 75 Inventory Transactions Inventory Status Forecasts Customer Orders Master Production Schedule Engineering Changes MRP: (Explosion Offsets, Nets) Bill of Materials Exception Report & Schedules

76 What is the Purpose and Logic of MRP? Coordination of Production and Inventory in large, multi-stage production systems Used for - Scheduling & re-scheduling - Capacity Planning - Supplier coordination (internal & external) Timely dissemination of information Time-phased production & procurement - with lead time offsets & BOM explosions Independent vs. Dependent demand Requires centralized information system; hence ERP Organizes large complex production and delivery coordination requirements 76

77 Criticisms of MRP Deterministic Model Push system poor data ==> GIGO Self-fulfilling lead times Difficult/costly ot install & maintain Centralized command & control mindset

78 Cisco s End-to-End Integration for its Fulfillment Supply Chain 78 Customers New product development on-line with supply base Technology Supply Chain Design: Innovation through Acquisition Order info flows direct to Cisco and suppliers Cisco Single enterprise information system Dynamic replenishment, direct fulfillment, merge in transit Customer orders through Cisco Connection online Finished Product flows direct to customer via logistics supplier Contract Manufacturers Component Suppliers & Distributors Basic Design Principle: Arm s length Relationship with Fulfillment Chain Partners

79 Cisco s Strategy for Technology Supply Chain Design 1.Integrate technology around the router to be a communications network provider. 2. Leverage acquired technology with - sales muscle and reach - end-to-end IT - outsourced manufacturing - market growth 3. Leverage venture capital to supply R&D Basic Design Principle: Acquisition Relationship with Technology Chain Partners 79

80 Volatility Amplification in the Supply Chain: The Bullwhip Effect 80 Customer Retailer Distributor Factory Tier 1 Supplier Equipment Information lags Delivery lags Over- and underordering Misperceptions of feedback Lumpiness in ordering Chain accumulations SOLUTIONS: Countercyclical Markets Countercyclical Technologies Collaborative channel mgmt. (Cincinnati Milacron & Boeing)

81 Supply Chain Volatility Amplification: Machine Tools at the tip of the Bullwhip We are experiencing a 100-year flood. J. Chambers, 4/16/01 See "Upstream Volatility in the Supply Chain: The Machine Tool Industry as a Case Study," E. Anderson, C. Fine & G. Parker Production and Operations Management, Vol. 9, No. 3, Fall 2000, pp

82 LESSONS FROM A FRUIT FLY: CISCO SYSTEMS KNOW YOUR LOCATION IN THE VALUE CHAIN 2. UNDERSTAND THE DYNAMICS OF VALUE CHAIN FLUCTUATIONS 3. THINK CAREFULLY ABOUT THE ROLE OF VERTICAL COLLABORATIVE RELATIONSHIPS 4. INFORMATION AND LOGISTICS SPEED DO NOT REPEAL BUSINESS CYCLES OR THE BULLWHIP.

83 Class #8: Basic Concepts in Queueing System Performance = f(system parameters) Output/throughput rate Inventory Level/Queue Size/ ( λ ) ( ) Line length Waiting Time/Cycle Time ( W ) Capacity or Server utilization ( ρ ) Probability that Queue is full ( P full ) Arrival rate ( λ ) Service rate ( µ ) Service time ( M ) Number of servers ( S ) Queue/Buffer capacity ( R ) Capacity or Server utilization ( ρ ) Number of Service classes ( K )

84 Fish Processing Example Input Rate (Tons per month) Time (Months) 600

85 Fish Processing Example (con t) λ = 3600 then 4800 then 600 µ=3000, R= unlimited Freezer Inventory (Exhibit 3) Average Inventory = 1/3 x (2400/2) + 1/3 x ( )/2 + 1/3 x (9600/2) = ( )/3 = tons

86 Fish Processing Example (con t) λ = 3600 then 4800 then 600 µ=3000, R= Freezer Inventory (Exhibit 4) Average Inventory = 1/3 x (2400/2) + 1/3 x /12 x (2400/2) + 1/4 x 0 = (1200/ / /12) = 1300 tons

87 Fish Processing Example (con t) λ = 3600 then 4800 then 600 µ=3300, R= Freezer Inventory (Exhibit 5) Average Inventory = 1/3 x (1200/2) +.8/12 x /12 x (2400) +.89/12 x (2400/2) = ( ) = 1049 tons

88 Fish Processing Example (con t) λ = 3600 then 4800 then Average Thruput Capacity Inventory per month Utilization µ=3000, R= 4000 tons 3000 (.63) 100% 88 µ=3000, R= tons 2400 (.63) 80% µ=3300, R= tons 2600 (.63) 79% Tradeoffs: Cost of processing capacity vs. Cost of Storage Capacity vs. Value of output (net of holding costs)

89 89 Basic Concepts in Queueing: Capacity Utilization in Stochastic Systems Suppose avg arrival rate = 1/minute Avg service rate = 1.33/minute (or avg service time = 45 seconds) time arrivals

90 90 Basic Concepts in Queueing: Capacity Utilization in Stochastic Systems System Performance = f(system parameters) Capacity or Server utilization ρ = λ / µ (arrival rate/service rate) ρ = λ (S x µ) ρ = λ x M

91 Basic Concepts in Queueing: Little s Law System Performance = f(system parameters) Conservation of Flows in Stochastic Systems L = λ x W Avg Length of the Queue = Arrival rate x Avg Waiting time 600 MBA s = 300/year x 2 years

92 92 Basic Concepts in Queueing: Nonlinearities in Congestion in Stochastic Systems If service times and interarrival times have exponential distributions, then L= ρ 2 /(1-ρ) Total wait in the queue W= ρ 2 /λ(1-ρ) W A A ρ B ρ 0 1 W B (Arrival Rate / Service Rate = ρ) = congestion

93 Basic Concepts in Queueing: Nonlinearities in Congestion in Stochastic Systems 93 System Performance = f(system parameters) With exponential (λ) interarrivals, and service times with mean = M and std dev = σ, Then W= λ (M 2 +σ 2 ) 2(1-ρ)

94 Volatility Amplification in the Supply Chain: The Bullwhip Effect Order Order Order Info Info Info Retailer Wholesaler Distributor Factory How does production control work in the Beer Game? Information lags Delivery lags SOLUTIONS: Over- and underordering Countercyclical Markets Misperceptions of feedback Countercyclical Technologies Collaborative channel mgmt. Lumpiness in ordering (Cincinnati Milacron & Boeing) Chain accumulations

95 Management of Queues The Physics of Waiting Lines Number and type of servers Waiting time, service time, and system time Queue discipline Number of people in queue System utilization Total time in the queue W A 0 A ρ B ρ W B Cost Capacity Cost Delay Cost Total Cost Congestion 1 0 Congestion (Arrival Rate / Service Rate) (Arrival Rate / Service Rate) 1

96 Management of Queues 96 The Psychology of Waiting Lines SERVERS CUSTOMERS WAITING LINE Propositions 1. Unoccupied time feels longer than occupied time 2. Process waits feel longer than in process waits 3. Anxiety makes waits seem longer 4. Uncertain waits seem longer than known, finite waits 5. Unexplained waits are longer than explained 6. Unfair waits are longer than equitable waits 7. The more valuable the service, the longer the customer will wait 8. Solo waits feel longer than group waits

97 15.760: National Cranberry Case Admin: Webvan case; UHS case 2. What are the sources of variability in the NCC case? 3. What are the problems NCC is experiencing that should be addressed? 4. Describe the Process Flow Diagram. 5. Assess possible options for relieving truck waiting. 6. How would you assess converting some dry bins to wet? 7. How would you assess whether you can begin at 8 am? 8. How would you assess labor cost impacts? 9. How would you deal with the distribution of wet/dry and volume over the days of the season? 10. Can you eliminate/reduce demand peaks?

98 National Cranberry Process Flow Diagram weighed/ graded tested/ sampled Unload 5-10 min/truck Freeze dry bbls =4000 bbls both bbls =2000 bbls wet x400 bbls = 1200 bbls 1500 bbls/hr dry destone wet destone dechaff 1500 bbls/hr dechaff 2 x 1500 bbls/hr 2 x 1500 bbls/hr 3 x 400 bbls/hr separate dry 3 x 200 bbls/hr Bulk Bins 800 bbl/hr Bulk Truck 2000 bbl/hr Bag 667 bbl/hr Freeze

99 Wet Cranberry Inventory Buildup Assume: buildup x 70% wet =12600 bbl/day 12600/12=1050 bbls/hr; Plant begins operations at 11:00; Drying 600 bbl/hr Truck waiting = hrs x (4600/2)/75 = 511 hours x4 450x8 3200/1050 = trucks begin waiting at 10:03 am -600/hr 7:00 9:00 11:00 13:00 15:00 17:00 19:00 21:00 23:00 1:00 3: /600 = 7.67hrs = 2:40 am No more trucks Plant is empty after 7800/600 = 13 hours after 19:00 or 8 am the next morning Total run time = 12600/600 = 21 hours

100 Wet Cranberry Inventory Buildup Assume: buildup x 70% wet =12600 bbl/day 12600/12=1050 bbls/hr; Plant begins operations at 7:00; Drying 600 bbl/hr Truck waiting = x12 = hrs x (2200/2)/75 = 127 hours /hr 3200/450 = 7.1 = trucks begin waiting at 14:06 pm 2200/600 = 3.67hrs = 20:40 pm No more trucks 7:00 9:00 11:00 13:00 15:00 17:00 19:00 21:00 23:00 1:00 3:00 Plant is empty after 5400/600 = 9 hours after 19:00 or 4 am the next morning Total run time = 12600/600 = 21 hours

101 Wet Cranberry Inventory Buildup Assume: buildup x 70% wet =12600 bbl/day 12600/12=1050 bbls/hr; Plant begins operations at 7:00; Drying 800 bbl/hr Truck waiting = 0!! x12 =3000 No truck buildup 3000/800 = 3.75hrs = 20:45 pm Processing complete 7:00 9:00 11:00 13:00 15:00 17:00 19:00 21:00 23:00 1:00 3:00 Plant is empty after 3000/800 = 3.75 hours after 19:00 or 20;45 the next morning Total run time = 12600/800 = hours; dry berry processing drops to 400/hr

102 Basic Concepts in Queueing System Performance = f(system parameters) Output/throughput rate Inventory Level/Queue Size/ ( λ ) ( ) Line length Waiting Time/Cycle Time ( W ) Capacity or Server utilization ( ρ ) Probability that Queue is full ( P full ) Arrival rate ( λ ) Service rate ( µ ) Service time ( M ) Number of servers ( S ) Queue/Buffer capacity ( R ) Capacity or Server utilization ( ρ ) Number of Service classes ( K )

103 Kiwanee Dumpers:Capacity Analysis Busy Day: Arrival rate = 18,000 bbl/day = 1500bbl/hr = 20 trucks/hr L= ρ 2 /(1-ρ) W= ρ 2 /λ(1-ρ)

104 104 Basic Concepts in Queueing: Nonlinearities in Congestion in Stochastic Systems If service times and interarrival times have exponential distributions, then L= ρ 2 /(1-ρ) Total wait in the queue W= ρ 2 /λ(1-ρ) W A A ρ B ρ 0 1 W B (Arrival Rate / Service Rate = ρ) = congestion

105 Management of Queues The Physics of Waiting Lines Number and type of servers Waiting time, service time, and system time Queue discipline Number of people in queue System utilization Total time in the queue W A A ρ B ρ W B Capacity Cost Delay Cost Total Cost 0 Congestion 1 0 Congestion (Arrival Rate / Service Rate) (Arrival Rate / Service Rate) Cost 1

106 Management of Queues 106 The Psychology of Waiting Lines SERVERS CUSTOMERS WAITING LINE Propositions 1. Unoccupied time feels longer than occupied time 2. Process waits feel longer than in process waits 3. Anxiety makes waits seem longer 4. Uncertain waits seem longer than known, finite waits 5. Unexplained waits are longer than explained 6. Unfair waits are longer than equitable waits 7. The more valuable the service, the longer the customer will wait 8. Solo waits feel longer than group waits