WHY NETWORK PLANNING? _ Find the right balance between inventory, transportation and manufacturing costs, _ Match supply and demand under uncertainty by positioning and managing inventory effectively, _ Utilize resources effectively by sourcing products from the most appropriate manufacturing facility TIGA LANGKAH HIRARKI _ Network design Number, locations and size of manufacturing plants and warehouses Assignment of retail outlets to warehouses Major sourcing decisions Typical planning horizon is a few years. _ Inventory positioning: Identifyingstocking points Selecting facilities that will produce to stock and thus keep inventory Facilities that will produce to order and hence keep no inventory Related to the inventory management strategies _ Resource allocation: Determine whether production and packaging of different products is done at the right facility What should be the plants sourcing strategies? How much capacity each plant should have to meet seasonal demand? NETWORK DESIGN _ Physical configuration and infrastructure of the supply chain. _ A strategic decision with long-lasting effects on the firm. _ Decisions relating to plant and warehouse location as well as distribution and sourcing REEVALUASI INFRASTRUKTUR _ Changes in: demand patterns product mix production processes sourcing strategies cost of running facilities. _ Mergers and acquisitions may mandate the integration of different logistics networks KEY STRATEGIC DECISIONS _ Determining the appropriate number of facilities such as plants and warehouses. _ Determining the location of each facility. _ Determining the size of each facility. _ Allocating space for products in each facility. _ Determining sourcing requirements. _ Determining distribution strategies, i.e., the allocation of customers to warehouse OBJECTIVE AND TRADE OFFS
Transcript
WHY NETWORK PLANNING?_ Find the right balance between inventory,transportation and manufacturing costs,_ Match supply and demand under uncertaintyby positioning and managing inventoryeffectively,_ Utilize resources effectively by sourcingproducts from the most appropriatemanufacturing facility
TIGA LANGKAH HIRARKI_ Network designNumber, locations and size of manufacturing plants and warehousesAssignment of retail outlets to warehousesMajor sourcing decisionsTypical planning horizon is a few years._ Inventory positioning: Identifyingstocking pointsSelecting facilities that will produce to stock and thus keep inventoryFacilities that will produce to order and hence keep no inventoryRelated to the inventory management strategies_ Resource allocation:Determine whether production and packaging of different products isdone at the right facilityWhat should be the plants sourcing strategies?How much capacity each plant should have to meet seasonaldemand?
NETWORK DESIGN_ Physical configuration and infrastructure ofthe supply chain._ A strategic decision with long-lasting effectson the firm._ Decisions relating to plant and warehouse
location as well as distribution and sourcing
REEVALUASI INFRASTRUKTUR_ Changes in:demand patternsproduct mixproduction processessourcing strategiescost of running facilities._ Mergers and acquisitions may mandate theintegration of different logistics networks
KEY STRATEGIC DECISIONS_ Determining the appropriate number offacilities such as plants and warehouses._ Determining the location of each facility._ Determining the size of each facility._ Allocating space for products in each facility._ Determining sourcing requirements._ Determining distribution strategies, i.e., theallocation of customers to warehouse
OBJECTIVE AND TRADE OFFS_ Objective: Design or reconfigure the logistics network inorder to minimize annual system-wide cost subject to avariety of service level requirements_ Increasing the number of warehouses typically yields:An improvement in service level due to the reduction in average traveltime to the customersAn increase in inventory costs due to increased safety stocks requiredto protect each warehouse against uncertainties in customerdemands.An increase in overhead and setup costsA reduction in outbound transportation costs: transportation costs
from the warehouses to the customersAn increase in inbound transportation costs: transportation costs fromthe suppliers and/or manufacturers to the warehouses.
DATA COLLECTIONS_ Locations of customers, retailers, existing warehouses anddistribution centers, manufacturing facilities, and suppliers._ All products, including volumes, and special transport modes(e.g., refrigerated)._ Annual demand for each product by customer location._ Transportation rates by mode._ Warehousing costs, including labor, inventory carryingcharges, and fixed operating costs._ Shipment sizes and frequencies for customer delivery._ Order processing costs._ Customer service requirements and goals._ Production and sourcing costs and capacities
DATA AGREGATION_ Customer ZoneAggregate using a grid network or other clustering technique forthose in close proximity.Replace all customers within a single cluster by a single customerlocated at the center of the clusterFive-digit or three-digit zip code based clustering._ Product GroupsDistribution pattern▪ Products picked up at the same source and destined to the same customers▪ Logistics characteristics like weight and volume.Product type▪ product models or style differing only in the type of packaging.
REPLACING ORIGINAL DATA WITH AGREGAT DATA
_ Technology exists to solve the logisticsnetwork design problem with the originaldata_ Data aggregation still useful because forecastdemand is significantly more accurate at theaggregated level_ Aggregating customers into about 150-200zones usually results in no more than a 1percent error in the estimation of totaltransportation costs
GENERAL RULE FOR AGREGATION_ Aggregate demand points into at least 200 zonesHolds for cases where customers are classified into classesaccording to their service levels or frequency of delivery_ Make sure each zone has approximately an equalamount of total demandZones may be of different geographic sizes._ Place aggregated points at the center of the zone_ Aggregate products into 20 to 50 product groups
AGGREGASI CUST BASED ON 3 DIGIT ZIP CODESTotal Cost:$5,796,000Total Customers: 18,000Total Cost:$5,793,000Total Customers: 800Cost Difference < 0.05%
TRANSPORTATION RATE_ Rates are almost linear with distance but notwith volume
_ Differences between internal rate andexternal rate
INTERNAL TRANSPORTATION RATE_ For company-owned trucks_ Data Required:Annual costs per truckAnnual mileage per truckAnnual amount deliveredTruck’s effective capacity_ Calculate cost per mile per SKU.
EXTERNAL TRANSPORTATION RATETWO MODE TRANSPORTATION_ Truckload, TL_ Country sub-divided into zones. One zone/stateexcept for:Big states, such as Florida or New York (two zones)_ Zone-to-zone costs provides cost per mile pertruckload between any two zones.TL cost from Chicago to Boston =Illinois-Massachusetts cost per mile X Chicago-Bostondistance_ TL cost structure is not symmetric_ Less-Than-Truckload, LTL_ Class ratesstandard rates for almost all products or commodities shipped.Classification tariff system that gives each shipment a rating or a class.Factors involved in determining a product’s specific class include:▪ product density, ease or difficulty of handling and transporting, and liability fordamage.After establishing rating, identify rate basis number.▪ Approximate distance between the load’s origin and destination.With the two, determine the specific rate per hundred pounds(hundred weight, or cwt) from a carrier tariff table (i.e., a freight ratetable).
_ Exception rates provides less expensive rates_ Commodity rates are specialized commodity-specific rates
SMC3’Z CZARLITE_ Engine to find rates in fragmented LTL industry_ Nationwide LTL zip code-based rate system._ Offers a market-based price list derived fromstudies of LTL pricing on a regional, interregional,and national basis._ A fair pricing system_ Often used as a base for negotiating LTL contractsbetween shippers, carriers, and third-party logisticsproviders
FIGURE 3-7: Transportation rates for shipping 4,000 lb
MILEAGE ESTIMATION_ Estimate lona and lata, the longitude andlatitude of point a (and similarly for point b)
CIRCUIT FACTOR, P_ Equations underestimate the actual road
distance._ Multiply Dab by ρ._ Typical values:ρ = 1.3 in metropolitan areasρ = 1.14 for the continental United States
CHICAGO BOSTON DISTANCE_ lonChicago = -87.65_ latChicago = 41.85_ lonBoston = -71.06_ lonBoston = 42.36_ DChicago, Boston = 855 miles_ Multiply by circuity factor = 1.14_ Estimated road distance = 974 miles_ Actual road distance = 965 miles_ GIS systems provide more accuracy_ Slows down systems_ Above approximation good enough!
WAREHOUSE COST_ Handling costsLabor and utility costsProportional to annual flow through the warehouse._ Fixed costsAll cost components not proportional to the amount offlowTypically proportional to warehouse size (capacity) but in anonlinear way._ Storage costsInventory holding costsProportional to average positive inventory levels.
DETERMINING FIXED COST
FIGURE 3-8: Warehouse fixed costs as a function of thewarehouse capacity
DETERMINING STORAGE COST_ Multiply inventory turnover by holding cost_ Inventory Turnover =Annual Sales / Average Inventory Level
WAREHOUSE CAPACITY_ Estimation of actual space required_ Average inventory level =Annual flow through warehouse/Inventory turnover ratio_ Space requirement for item = 2*Average Inventory Level_ Multiply by factor to account foraccess and handlingaisles,picking, sorting and processing facilitiesAGVs_ Typical factor value = 3
WH CAP. EXAMPLE_ Annual flow = 1,000 units_ Inventory turnover ratio = 10.0_ Average inventory level = 100 unitsAssume each unit takes _ 10 sqft. of space_ Required space for products = 2,000 sqft._ Total space required for the warehouse isabout 6,000 square feet
POTENTIAL LOCATIONS_ Geographical and infrastructure conditions._ Natural resources and labor availability._ Local industry and tax regulations._ Public interest._ Not many will qualify based on all the aboveConditions
SERVICE LEVEL REQ_ Specify a maximum distance between eachcustomer and the warehouse serving it_ Proportion of customers whose distance to
their assigned warehouse is no more than agiven distance95% of customers be situated within 200 miles ofthe warehouses serving themAppropriate for rural or isolated areas
FUTURE DEMAND_ Strategic decisions have to be valid for 3-5years_ Consider scenario approach and net presentvalues to factor in expected future demandover planning horizon
NUMBER OF WH
INDUSTRY BENCHMARKNUMBER OF DC
MODEL VALIDATION_ Reconstruct the existing network configuration using themodel and collected data_ Compare the output of the model to existing data_ Compare to the company’s accounting informationOften the best way to identify errors in the data, problematicassumptions, modeling flaws.
_ Make local or small changes in the network configuration tosee how the system estimates impact on costs and servicelevels.Positing a variety of what-if questions._ Answer the following questions:Does the model make sense?Are the data consistent?Can the model results be fully explained?Did you perform sensitivity analysis?
TEKNIK SOLUSI_ Mathematical optimization techniques:1. Exact algorithms: find optimal solutions2. Heuristics: find “good” solutions, not necessarilyoptimal_ Simulation models: provide a mechanism toevaluate specified design alternatives created bythe designer.
EXAMPLE_ Single product_ Two plants p1 and p2Plant p2 has an annual capacity of 60,000 units._ The two plants have the same production costs._ There are two warehouses w1 and w2 with identicalwarehouse handling costs._ There are three markets areas c1,c2 and c3 withdemands of 50,000, 100,000 and 50,000,respectively.
UNIT DISTRIBUTION COST
THE OPTIMIZATION MODELThe problem described earlier can be framed as thefollowing linear programming problem.Let_ x(p1,w1), x(p1,w2), x(p2,w1) and x(p2,w2) be the flows fromthe plants to the warehouses._ x(w1,c1), x(w1,c2), x(w1,c3) be the flows from the warehouse
w1 to customer zones c1, c2 and c3._ x(w2,c1), x(w2,c2), x(w2,c3) be the flows from warehouse w2to customer zones c1, c2 and c3
The problem we want to solve is:min 0x(p1,w1) + 5x(p1,w2) + 4x(p2,w1)+ 2x(p2,w2) + 3x(w1,c1) + 4x(w1,c2)+ 5x(w1,c3) + 2x(w2,c1) + 2x(w2,c3)
subject to the following constraints:x(p2,w1) + x(p2,w2) 60000x(p1,w1) + x(p2,w1) = x(w1,c1) + x(w1,c2) + x(w1,c3)x(p1,w2) + x(p2,w2) = x(w2,c1) + x(w2,c2) + x(w2,c3)x(w1,c1) + x(w2,c1) = 50000x(w1,c2) + x(w2,c2) = 100000x(w1,c3) + x(w2,c3) = 50000all flows greater than or equal to zero.
OPTIMAL SOLUSI
MODEL SIMULASI_ Useful for a given design and a micro-levelanalysis. Examine:Individual ordering pattern.Specific inventory policies.Inventory movements inside the warehouse._ Not an optimization model_ Can only consider very few alternate models
WHICH ONE TO USE?_ Use mathematical optimization for staticanalysis_ Use a 2-step approach when dynamics insystem has to be analyzed:Use an optimization model to generate a number
of least-cost solutions at the macro level, takinginto account the most important costcomponents.Use a simulation model to evaluate the solutionsgenerated in the first phase.
DSS FOR NETWORK DESIGN_ Flexibility to incorporate a large set of preexisting networkcharacteristics_ Other Factors:Customer-specific service level requirements.Existing warehouses kept openExpansion of existing warehouses.Specific flow patterns maintainedWarehouse-to-warehouse flow possibleProduction and Bill of materials details may be important_ RobustnessRelative quality of the solution independent of specific environment,data variability or specific settings
3.3. INVENTORY POSITIONING_ Multi-facility supply chain that belongs to a single firm_ Manage inventory so as to reduce system wide cost_ Consider the interaction of the various facilities and theimpact of this interaction on the inventory policy of eachfacility_ Ways to manage:Wait for specific orders to arrive before starting to manufacture them[make-to-order facility]Otherwise, decide on where to keep safety stock?Which facilities should produce to stock and which should produce toorder?_ Assume -SI: amount of time between when an order is placed untilthe facility receives a shipment (Incoming Service Time)S: Committed Service Timemade by the facility to its owncustomers.
T: Processing Time at the facility._ Net Lead Time = SI + T - S_ Safety stock at the facility:SI T Szh SI T SReducing committed service time from facility 2to facility 1 impacts required inventory at bothfacilitiesInventory at facility 1 is reducedInventory at facility 2 is increasedOverall objective is to choose:the committed service time at each facilitythe location and amount of inventoryminimize total or system wide safety stock cost._ Large contract manufacturer of circuit boards and other hightech parts._ About 27,000 high value products with short life cycles_ Fierce competition => Low customer promise times< Manufacturing Lead Times_ High inventory of SKUs based on long-term forecasts =>Classic PUSH STRATEGYHigh shortagesHuge risk_ PULL STRATEGY not feasible because of long lead times_ OBJECTIVES:Reduce inventory and financial risksProvide customers with competitive response times._ ACHIEVE THE FOLLOWING:Determining the optimal location of inventory across the various stagesCalculating the optimal quantity of safety stock for each component at eachstage_ Hybrid strategy of Push and PullPush Stages produce to stock where the company keeps safety stockPull stages keep no stock at all._ Challenge:
Identify the location where the strategy switched from Push-based to PullbasedIdentify the Push-Pull boundary_ Benefits:For same lead times, safety stock reduced by 40 to 60%Company could cut lead times to customers by 50% and still reduce safetystocks by 30%FIGURE 3-11: How to read the diagrams_ If Montgomery facility reduces committed lead time to 13daysassembly facility does not need any inventory of finished goodsAny customer order will trigger an order for parts 2 and 3.▪ Part 2 will be available immediately, since it is held in inventory▪ Part 3 will be available in 15 days▪ 13 days committed response time by the manufacturing facility▪ 2 days transportation lead time.Another 15 days to process the order at the assembly facilityOrder is delivered within the committed service time._ Assembly facility produces to order, i.e., a Pull basedstrategy_ Montgomery facility keeps inventory and hence is managedwith a Push or Make-to-Stock strategy.FIGURE 3-12: Current safety stock locationFIGURE 3-13: Optimized safety stockFIGURE 3-14: Optimized safety stock with reduced lead timeFIGURE 3-15: Current supply chainFIGURE 3-16: Optimized supply chain_ Identifying the Push-Pull boundary_ Taking advantage of the risk pooling conceptDemand for components used by a number of finishedproducts has smaller variability and uncertainty than thatof the finished goods._ Replacing traditional supply chain strategies that
are typically referred to as sequential, or local,optimization by a globally optimized supply chainstrategy.FIGURE 3-17: Trade-off between quoted lead time and safety stock_ For the same lead time, cost is reducedsignificantly_ For the same cost, lead time is reducedsignificantly_ Trade-off curve has jumps in various placesRepresents situations in which the location ofthe Push-Pull boundary changesSignificant cost savings are achieved._ Prevalent strategy for many companies:try to keep as much inventory close to the customershold some inventory at every locationhold as much raw material as possible._ This typically yields leads to:Low inventory turnsInconsistent service levels across locations and products,andThe need to expedite shipments, with resulting increasedtransportation costs_ Consider a two-tier supply chainItems shipped from manufacturing facilities to primarywarehousesFrom there, they are shipped to secondary warehousesand finally to retail outlets_ How to optimally position inventory in the supplychain?Should every SKU be positioned both at the primary andsecondary warehouses?, ORSome SKU be positioned only at the primary while othersonly at the secondary?FIGURE 3-18: Sample plot of each SKU by volume and demand
_ High variability - low volume products_ Low variability - high volume products, and_ Low variability - low volume products._ High variability low volume productsInventory risk the main challenge forPosition them mainly at the primary warehouses▪ demand from many retail outlets can be aggregated reducinginventory costs._ Low variability high volume productsPosition close to the retail outlets at the secondarywarehousesShip fully loaded tracks as close as possible to thecustomers reducing transportation costs._ Low variability low volume productsRequire more analysis since other characteristics areimportant, such as profit margins, etc._ Supply chain master planningThe process of coordinating and allocatingproduction, and distribution strategies andresources to maximize profit or minimize systemwidecost_ Process takes into account:interaction between the various levels of the supply chainidentifies a strategy that maximizes supply chainperformance_ Facility locations: plants, distribution centers and demandpoints_ Transportation resources including internal fleet andcommon carriers_ Products and product information_ Production line information such as min lot size, capacity,costs, etc.
_ Warehouse capacities and other information such as certaintechnology (refrigerators) that a specific warehouse has andhence can store certain products_ Demand forecast by location, product and time._ Sourcing Strategies:where should each product be produced duringthe planning horizon, OR_ Supply Chain Master Plan:production quantities, shipment size and storagerequirements by product, location and timeperiod.FIGURE 3-19: The extended supply chain: from manufacturing to order fulfillment_ Will leased warehouse space alleviate capacity problems?_ When and where should the inventory for seasonal or promotionaldemand be built and stored?_ Can capacity problems be alleviated by re-arranging warehouseterritories?_ What impact do changes in the forecast have on the supply chain?_ What will be the impact of running overtime at the plants or outsourcingsourcing production?_ What plant should replenish each warehouse?_ Should the firm ship by sea or by air. Shipping by sea implies longlead times and therefore requires high inventory levels. On theother hand, using air carriers reduces lead times and henceinventory levels but significantly increases transportation cost._ Should we rebalance inventory between warehouses or replenishfrom the plants to meet unexpected regional changes in demand?Network Design Inventory Positioningand ManagementResource AllocationDecision focus Infrastructure Safety stock Production DistributionPlanning Horizon Years Months MonthsAggregation Level Family Item Classes
Frequency Yearly Monthly/Weekly Monthly/WeeklyROI High Medium MediumImplementation Very Short Short ShortUsers Very Few Few Few_ Optimizing supply chain performance is difficultconflicting objectivesdemand and supply uncertaintiessupply chain dynamics._ Through network planning, firms can globally optimizesupply chain performanceCombines network design, inventory positioning andresource allocationConsider the entire network▪ account production▪ Warehousing▪ transportation inventory costs▪ service level requirements._ Demonstrate applicability of risk pooling andpostponement, EOQ modeling, andinventory sizing to improve customer servicein make-to-order job shop setting_ Demonstrates value from getting and lookingat data_ Background and context_ Why are lead times long?_ How might they be reduced?What are the _ costs? benefits?Stephen C. Graves Copyright 2003All Rights Reserved_ Make-to-order job shop operation_ 600 SKU’s made from 4” sheet bar (4 alloys)_ Goal to reduce 7-week customer lead times_ Expediting is ad hoc scheduling rule_ Six months of inventory_ Manufacturing cycle time is 2 – 3 weeks_ Limited dataStephen C. Graves Copyright 2003All Rights Reserved4” Bar 1/4” Plate 1/8” Plate 0.015” Sheet TubingSheet Bar Roll Clean Anneal(forged ingot)Repeat
0 n 3Finish(cut, weld, etc.)Production Order #1Production Order #2 Production Order #3Production OrdersStephen C. Graves Copyright 2003All Rights Reserved_ From sales order to process order takes 2weeks_ Typical order requires multiple processorders, each 2 – 3 weeks_ Expediting as scheduling rule_ Self fulfilling prophecy?Stephen C. Graves Copyright 2003All Rights Reserved_ New accounts and new business_ Protect current business from switching tosubstitutes or Chinese competitorPossibly _ less inventory_ Better planning and better customer service_ Savings captured by customers?Stephen C. Graves Copyright 2003All Rights Reserved_ Hold intermediate inventoryHow would this help?How much? Where?Eliminate _ paper-work delays_ Reduce cycle time for each process orderHow? What cost?Stephen C. Graves Copyright 2003All Rights Reserved2 22KB KF hBDB hFDFCost T TTK KStephen C. Graves Copyright 2003All Rights Reserved**2 400 4000.02 years
.06 100 526000 .06 125 183000B FB B F FTh D h DT_ Characterize demand by possibleintermediate for each of two alloys_ Pick stocking points based on risk poolingbenefits, lead time reduction, volume_ Determine inventory requirements based oninventory model, e. g. base stockStephen C. Graves Copyright 2003All Rights ReservedPopularity Material Gauge - Description Jan Feb Mar Apr May Jun Jul Aug Sep Total Cum %1 1011 0.002 Foil 618 1,079 1,215 1,188 1,020 290 1,590 849 1,017 8,866 22%2 1004 0.015 Sheet 68 611 1,263 167 1,917 803 321 377 404 5,931 37%3 1003 0.005 Sheet 263 576 584 812 617 969 572 359 909 5,661 50%4 1029 0.500 Disk - 10" dia 275 0 353 0 581 0 530 414 1,017 3,170 58%5 1009 0.030 Sheet 0 122 614 275 422 360 686 246 177 2,902 65%6 1008 0.040 Sheet 321 101 191 486 8 98 263 176 690 2,334 71%7 1002 0.010 Sheet 20 56 287 179 41 204 560 143 276 1,766 76%8 1014 0.250 Plate 6 12 0 770 0 752 0 0 174 1,714 80%9 1007 0.060 Plate 0 146 32 117 129 414 581 26 191 1,636 84%10 1012 0.125 Plate 228 8 32 90 432 17 8 0 450 1,265 87%11 1013 0.150 Plate 1,100 0 0 0 0 35 0 0 0 1,135 90%12 1028 0.500 Ring - 10" OD x 8.5" ID 0 189 0 48 293 93 0 0 174 797 92%1999 Invoiced Sales - Pounds per month13 1010 0.020 Sheet 0 54 102 183 45 54 126 92 119 775 94%14 1017 0.750 Tube - 3/4" 0 0 0 8 12 558 0 0 12 590 95%15 1015 0.375 Plate 0 0 0 0 0 0 375 0 0 375 96%
