A. I5 ystems for I Inventory S ndependent Den Outline 3 rae an Fixed-Time Period Models, 599 aa tantary 3" Fixed-Time Period Model Purposes of Inventor. ce a Inventory Costs, 584 Independent versus De Inventory Systems, 585 Classifying Models Fixed-Order Quantity Models. 587 = Fixed-Order Quantity Mode! with Usage Special-Purpose Models, 601 Miscellaneous Systems and Three Simple Inventory ABC Inventory Planning Inventory Accuracy anid Inventory Control in Servi pendent Demand, S85 During Production Time x Establishing Safety Stock Levels Fixed-Order Quantity Model with Specified Service Level ory Fixed-Time Period Models Res Materials (P Models) ea eee Inventar Fh ‘Work in Process Safety Stock Inkpen and Dependent Serice Level and Fixed-Order Quantity Model: 7 Oven) ge ee www Links gzJENERAL MOTORS CORP. iS OVERHAULING THE WAY IT gets vehicles to dealers in an effort to deliver the exaetcar that a customer wants in less than 24 hours. Under the new program, which will affect all GM brands, large inventories of ‘cars and light trucks will be kept in regional distribution centers across the US. rather © than being parked on dealer lots, according to people close tothe situation. That way, dealers who don’t have a particular vehicle in stock will be able to receive it from a regional center in less than a da) GM estimates that 35 perceat of auto customers don’t find the exact car they vant on dealer lots and are forced to make a compromise, another 1 percent switch dealerships; and 11 percent i general kind of car they want. ‘The plan at GM is to greatly expand @ Cadillac Division called Custom Xpress which began two years ago with a 32 eso liminte bugs om$82 fentory Managing the Supply “hain Section Four Question: Why would GM want to make this type of change in its di This is not an easy question to answer. It comes down to questions of ' Which would be better, a bunch of little inventories scattered all over: dealerships, or a few larger inventories from which customers can quic} cars? Mathematically, we can show that GM should be able to give better reduce its inventory investment with the distribution center approach, Are factors? Do custgmers want to see the different colors and options? Would cusig rather go to a big dealership with a full variety of cars on the lot? It will be i to see if GM's approach works. Consider this. The average cost of inventory across all manufac States is 30 to 35 percent of its value. For example, é $20 million, it costs the firm more than $6 million per year. Thebe costs are 4 obsolescence, insurance, opportunity costs, and so forth. If the amount of ir could be reduced to $10 million, for instance, the firm would save over $3 which goes directly to the bottom line. That is, the savings from reduced j shows as increased profit In this chapter, we present standard inventory models designed to help ment keep the cost down while still meeting production and customer service ments. Also included are special purpose models, such as price-break, as well as ABC technique. In addition, we discuss inventory accuracy and show applic the models in department stores and auto parts supply. There are conflicting views concerning the teaching of classical inventory. On one side. articles claiin that economic order quantity (EOQ) models are in The other side defends their use. We believe both sides are correct—from within own arenas. While you must be careful in their application, there certainly are tions in manufacturing where EOQ models can be successfully used, Justi manufacturing (JIT), for éxample, is based on the classical production-cons 'aventory model discussed here. Classical models are quite valid for the many sands of companies engaged in product and parts distribution. Concerning JIT and safety stocks. recall that JIT does have safety stock’ up as the size of containers and the number of containers between each production sequence. Further, all manufacturers that supply parts cannot have: JIT schedules. A manufacturer using a JIT system utilizing twice-per-day deli 2 supplier to the production line might be surprised that the supplier prod supplies a month's worth ata time using an EOQ formula! How we compute i requirements depends on many factors—all methods are vali given the of circumstances. Therefore. we should become. familar with them all. _ Curing in the U if a'firm carries an inventor: @ DEFINITION OF INVENTORY Inventory is the stock of any item Or resource used in ano nization system is the set of policies and controls that monitors levels of invent mines what levels should be maintained, when stock should be repl large orders should be. sity By convention, manufacturing inventory generally refers to ‘or become part of a firm's product output. inventory gene’ __ work in process. In services,Chapter 15 si Wentory Systems for Independent Demand ‘The basic purpose of inventory analysis in manufacturing and stockkeeping ser- vices is to specify (1) when items should be ordered and (2) how large the order to supply their needs for perhaps the entire year. This changes the “when” and “how many to order” to “when” and “how many to deliver.” PURPOSES OF INVENTORY ms All firms (including JIT operations) keep a supply-of inventory fps the follwene, reasons; 1. To maintain independence of ope’ allows that center flexibility in operat ‘waking each new production setup, this Inventory allows man: cumber ey. : i. in The po dee workstations is desirable on assembly ines 26 ees operations will naturally Yay ane isa vneushion of several parts within the Wor an for longer performance materials at a work center because there are costs for agement to reduce the rations. A supply of tions. For example, should be. Many firms are tending to enter into longer-term relationships with vendors, though an asset can be costly f hed x00 long by fine peiot GE explains cher‘ecion Four Managing the Sepply CHI" . id. If the demand for the “ jon in product demand. 1 2. To meet wariation Vile (though not necessarily econ q precisely, it may ince the demand. Usually, however, demand ig b jock must be maint 0 ki mn tee ‘ety or buffer stock must be maintained to absorb et nown, and a 8a ing. A stock of inve * allow flexibility in production scheduling. k 3 3 Ta ten auetion temo get the goods out. This causes tory re w Shih permit production planning for smoother flow and lower. a hich er Pine pradcton. High setup cot, fr example vor he < Sf a larger number of units once the setup has been made, 4. To provide a safeguard for variation in raw material delivery time, When, ic ordered from a vendor, delays can occur for a variety of reasons: a normaly in shipping time, a shortage of material at the vendor's plant causing bac tinexpected strike at the vendor's plant or at one of the shipping companies ‘ order, or a shipment of incorrect or defective material, / $. To take advantage of economic purchase-order size, There are costs to order: labor, phone calls, typing, postage, and so on. “Therefore, the larger the: . the fewer the number of orders that need be written. Also, shippis the larger the shipment, the lower the per-unit cost, For each of the preceding reasons (especially for items 3, 4, and 5), be inventory is costly and large amounts are generally undesirable. Long cycle caused by large amounts of inventory and are undesirable as well. @ INVENTORY COSTS sven 6 In making any decision that affects inventory size, the following costs considered. 1. Holding (or carrying) costs. This broad category includes the costs for facilities, handling, insurance, pilferage, breakage, obsolescence, depres and the opportunity cost of capital. Obviously, high holding costs tend to inventory levels and frequent replenishment. : ‘ 2. Setup (or production change) costs. To make each different product obtaining the necessary materials, arranging specific equipment setups, filling required papers, appropriately charging time and materials, and moving previous stock of material : If there were no costs or Joss of time in changing from one product many small lots would be produced. This would reduce inventory resulting savings in cost. One challenge today is to try to reduce these permit smaller lot sizes. (This is the goal of a JIT system.) x Ordering costs. These costs refer to the managerial and clerical Purchase or production order. Ordering costs include all the details, items and calculating order quantities. The costs associated wit! system needed to track orders are also jgeluded in ordering—== ‘tory Systems for Independemt Demand Inver Fae . a * from vendors or the size of lots submitted trom the combined effects of four maneg Sea" forthe minimum total cou resulting or ed short Oren vidual Costs: holding costs, setup costs. ordering may impact iNVENtOry cost, ibe Uming of these orders is a critical factor that i INDEPENDENT VERSUS DEPENDENT DEMAND. In inventory management, itis important to underst lependent demand. The reason is that entre meena eee ee cated on whether demand is derived from an end Briefly, the distinction between independent {n concept, dependent demand is a relatively straightforward computational prob- em. Needed quantities of a dependent-demand item are simply computed. based on the number needed in each higher-level item in which it is used. For example, if an automobile company plans on producing 500 cars per day, then obviously it will need 2.000 wheels and tires (plus spares). The nuinber of wheels and tires needed is dependent on the production levels and is not derived separately. The demand for cars. on the other hand, is independeni—it comes from many sources exterwal to the utomobile firm and is not a part of other products; it is unrelated to the demand for other products. To determine the quantities of independent items that must be produced, firms usually turn to their sales and market research departments. They use a variety of ‘cchnigues, including customer surveys, forecasting techniques, and economic and ‘vclgicl trends, as we discussed in Chapter 13 on forecasting, Because indepen. eot demand is uncertain, extra units must be carried in inventory. This chapter Independent and Dependent Demand : cl bresents models to determine how manly units need to be ordered, and how many Unis should be carried to provide a specified service level se emand) that the firm would like to satisfy immediately neo giived-Th Models ( ’ Far Managing the Supply Chain ime Period P Models) Seetion Classifying Models . ‘There are two general types of inventory em . as Thal the economic onder quantity. BDO. 0 =atS models (also referred t0 variouly ag ystem, periodic. erval system, fed. order an is that Bixed order quantity models are “event fixed-time period models are “time triggered.” That is, a fixed-order quan ties an order when the event of reaching a specified reorder level ama take place at any time, depending on the demand for the items In contrast, the fixed-time period model limited to placing orders at the « Termined time period: only the passage of time triggers the model To ave the fixed-order quantity model (which places an order when the re 'y drops to a predetermined order point, R). the inventory Continually monitored. Thus, the fixed-order quantity model is a pe hich requires that every time a withdrawal from inventory or an addition ty ‘bry is made, records must be updated to ensure that the reorder point has: teen reached. In a fixed-time period model counting takes place only at period. (We will discuss some variations of systems that combine features Some additional differences that tend to influence the choice of systems g see Exhibit 15.1): 7 inventor The fixed-time period model has a larger average inventory beeat also protect against stockout during the review period, 7; the fixed model has no review period. The fixed-order quantity model favors more expensive items because: inventory is lower. ‘The fixed-order quantity model is more appropriate for important it as critical repair parts because there is closer monitoring and 1 quicker response to potential stockout. : The fixed-order quantity model requires more time to maintain addition or withdrawal is logged. Exhibit 15.2 shows what occurs when each of the two models is put becomes an operating system. As we can see. the fixed-order quantity Fxed-Order Quantity and Fixed-Time Period Differences io Fixed-Order Quantity Model +.== tape on order quantities and reorder points. Procedurally, each time a unit is taken out of stock, the withdrawal is logged and the amount remaining in inventory is immediately compared to the reorder point. If it has dropped to this point, an order for Q items is placed. If it has not, the system remains in an idle state until the next withdrawal. In the fixed-time period system, a decision to place an order is made after the stock has been counted or reviewed. Whether an order is actually placed depends on the inventory position at that time. @ FIXED-ORDER QUANTITY MODELS Fixed-order quantity models attempt to determine the specific point, R, at which an order will be placed anid the size of that order, Q. The order point, R, is always a specified number of units. An order of size @ is placed when the inventory available Chapter 15 “Inventory Systems for Independent Demand ~~ (currently in stock and on order) reaches the point R. Inventory position is defined Inventory Position as the on-hand plus on-order minus back ordered quantities. The solution to a fixed- order quantity model may stipulate something like this: When the inventory position drops to 36, place an order for 57 more units. ‘The simplest models in this category occur when all aspects of the situation are known with certainty. If the annual demand for a product is 1,000 units, itis precisely 1,000—not 1,000 plus or minus 10 percent. The same is true for setup costs and a a Pa i Fixed-Order, Hee hae te Quantity System Berens ran oo | [ =a Ie | Idle state Waiting for demand | Waiting for demand | 1 ‘Demand occurs | Demand occurs Unit withdrawn {ah Unite withdrawn from Raeiestebey ok | inventory or back ordered ‘back ordered | ; ‘Compute inventory position Position = On hand + On order ~ Back order ue inventory position | c Smee 5 ‘On order ~ Back order | Yes Ase cH ia 5 a Comparison of Fixed-Order Quantity and Fixed-Time Period Reordering Inventory Systems——=—ai i Mu Section Four though the assumption of complete certainty i I for our coverage of inventory models. * the discussion about deriving the optimal order qua ‘on the following characteristics of the model. These assumptions are they represent a starting point and allow us to uSe a ple example, Demand for the product is constant and uniform thro ‘om ordering to receipt) is constant. holding costs. Al vides a good basi Exhibit 15.3 + Lead time (time fr Price per unit of product is constant. Inventory holding cost is based on average inventory, Ordering or setup costs are constant. All demands for the product will be satisfied. (INo back orders are The “sawtooth effect” relating Q and R in Exhibit 15.3 shows that when the tory position drops to point R, a reorder is placed. This order is received at the time period L, which does not vary in this model In constructing any inventory model. the first step is to develop a functional tionship between the variables of interest and the measure of effectiveness. In case, because we are concerned with cost. the following equation pertains: Total ., Annual Annual =, Annual annual cost purchase cost. ordering cost sholding cost or V4 c= oc - 25 42H Q Z Basic Fixed-Order Quantity Model Inventory ‘on handTotal annual cost ~ Demand (annual) = Cost per unit Quantity to be ordered : order quantity EOQ— Ai a amoui Setup cost or cost of placing am order R = Reorder point L = Lead time ‘Be Cost per unit of average inventory (Often, Percentage of the cost of the item, such as cent carrying cost.) holding cost is taken as a H = iC where i is the per st of each order, S), and (Q/2)H is the annual holding cost (the average ory, Q/2, times the cost per unit for holding and storage, H). These cost ships are graphed in Exhibit 15.4. he second step in model development isto find that order quantity Qu, at which | cost is a minimum. In Exhibit 15.4, the total cost is minimum at the point where be slope of the curve is zero. Using calculus, we take the derivative of total cost with ect to Q and set this equal to zero. For the basic model considered here, the alculations are Dy, @ TC = DOR Sane (45.2) Because this simple model assumes constant demand dnd lead time,no safety stock 's necessary, and the reorder point, R, is simply R=dL (15.3) where 7 d = Average daily demand (constant) L = Lead time in days (constant) Point mexample 15.1 Economic Order Quetiy and Reorder Poi nomic order quantity and the reorder point, Bh ; Annual demand (D) = 1,000 units ‘Average daily demand (d) = 1,000/365 Ordering cost (S) = $5 per ‘i Holding cost (H) = $1.25 pet unit per a Lead time (L) = 5 days Cost per unit (C) = $12.50 3 quantity should be ordered? Find the eco-———— ovine Fur Managing te SPRY Cid Eo a ‘ . jantity is +a ‘ solution ee oe : pce oon Vion 3 es ni is s oooeuanl tial R= db = ©) = 13.7 units 365 exanl icy is as follows: When the: wwentory- Rounding to the nearest unit, the inventory pol in’ ian position drops o 14, place an order for 89 more. 3 : operat The total annual cost will be a in a 2, 100 Sea re= pc +2s.+ Sn Give Q = . i 1,000, 89 = 1,000(12.50) ~ =5-(S) + (1.25) = $12,611.81 cost f Note that in this example. the purchase cost of the units was not d determine the order quantity and the reorder point because the Cost was €0 ¢ unrelated to order size. @ Wy § Z solut Fixed-Order Quantity Model with Usage During follows: Production Time Equation 15.1 assumed that the quantity ordered would be received in one lot s frequently this isnot the case. In many situations. production of an invent usage of that item take place simultaneously. This is particularly true where 0 of a production system acts asa supplier to another part. For example, while alumi This sta extrusions are being made to fill an order-for aluminum windows, the extrusions stock di cut and assembled before the entire extrusion order is completed. Also, ; Atl beginning to enter Jonger-term arrangements with suppliers. Under such supply single order may cover product or material needs over a six-month or year peria occupie y ; duced. denote a constant demand rate for some item going into production and let, production rate of tha process that uses the item, we may develop the Estah The d re TC = DC + as in aaa case » therefo Again differentiating with respect to Q and seiihg pies eaxetea Jor nao ena , 7 Kaupier 19° suseinory 2 = ==, xm Production taking pace pie, olin, mar Model with Usage aa during Production, ® Time example 15.2 Optimal Lot Size Product X is a standard item ina firm's ventory. Final assembly of the product is performed on an assembly line that is in ation every day. One component of product X (call it component X,) is produced another department. This department, when it produces X,, does so at the rate of units per day: The assembly line uses component X, at the rate of 40 units per day. Given the following data, what is the optimal lot size for production of component Daily usage rate (d) = 40 units Annual demand (D) = 10,000 (40 units x 250 working days) Daily production (p) = 100 units Cost for production setup ($) = $50 Annual holding cost (H) = $0.50 per unit Cost of component X,(C) = $7 each Lead time (L) = 7 days olution The optimal order quantity and the reorder point are calculated as _ PPS _p _ _ 2(10,000)50 100 ; on = a a DS0T e+ Too rageeteze os R = dL = 40(7) = 280 units his states that an order for 1,826 units of component X, should be placed when the tock drops to 280 units. S At 100 units per day, this run would take 18.26 days and provide a 45.65-day pply for the assembly line (1,826/40). Theoretically, the department would be cupied with other work for the 27.39 days when component X, is not being pro- ced. stablishing Safety ‘Stock Levels and known. In the majority cE ‘ee previous model assumed that demand was constant and ; { cases, though, demand is not constant but varies from day to day. Safety stock must‘eile out of ‘meet 95 percent , the first approach deals with the prob Is percent of the Sime) AB ach is concemed the how many units ae ue, and the secon’ Sxceeding & ¥al Short. ee aoe the probability criterion to determine Ke Probable APPT he reid deseribed in this chapter, we assume th , sioek is prety simple: Wane is normally distributed with a mean and a seman ove 8 Pe anber that this approach only considers the probab 2 AR not how many units we are short, To determine the p ‘te ume period, we can simply plot a normal distribution for d note where the amount we have on hand lies on the curve, ree take a few simple examples to illustrate this. Let’s say we expect demand be 100 units overthe next month, and we know thatthe standard deviation is 20 Trwe go into the month with just 100 units, we know that our probability of st st is 50 percent. Half of the months we would expect demand to be greater ‘00 units; half of the months, we would expect it to be less than 100 units. Taking thi irther, it we ordered amonth's worth of inventory of 100 units at a time and receive xt the beginning of the month, over the Jong run we would expect to run out aventory in 6 months of the year. If we found that running out this often was not acceptable, we would want to cart extra inventory-o reduce this risk of stocking out. One idea might be to carry an eit 20 units of inventory for the item. In this case. We woald still order a month's Wor f inventory at a time. but we would schedule delivery to arrive when we still have units remaining in inventory. This would give us that little cushion of safety reduce the probability of stocking out. If-the standard deviation associated with demand were 20 units, we would then be carrying one standard deviation safety stock. Looking at the Standard Normal Distribution (Appendix D), and movi cone standard deviation wo the right of the mean, gives a probability of 8413 ( table we get 3413 and weneed to add .5 to this), So, approximately 84 percent time we would not expect to stock out, and 16 percent of the time we would. \we order every month, we would expect to stock out approximately 2 months pe (16 X 12 = 1.92) It is comion for companies using this approach to set the probal stocking out at 95 percent. This would mean that we would carry about 1.64 deviations of safety stock, or 33 units (1.64 X 20 = 32.8) for our again. keep in mind that this does not mean that we would order 33 units € month, Rather, it means that we would still order a month's worth each mud Schedule the receipt so that we could expect to have 33 units in in order arrives. In this case, we would expect to stock-out approxim Per Year, or that sigck-outs would occur in 1 month of every 20 devia running out of of stocking out over expected demand anc The Service Level Approach Using an analogy, we will convey in using the probability approach to Sana eee p prdeng, Peds min emer Are you satisfied with the yes toler? ‘you prefer to know he. prediction it fe ‘tise wih a simplest of stow a To compt assume that deviation of we expect fc demanded ( ability that number of t While th nately, Rob Exhibit 1 units short model Q). l 100 units ar must multip of one unit numbers in deviation), because the during the j level is 100 If in the We would b °F 96.01 Again, Standard than 100°, 89.17 ber, = Chapter 1 lnventory Systems tor Independent erand re correct)? Wouldn’t you rather know whether this is a li Swi rsuling st asardos devin ands prob ec ot Mm ner eae, We are interested not only in whether we will be oS eee now) but also in how many units we will be short We are now ready to define service level: S¢ demanded that can-be supplied from stock curre jemand for an item is 1,000 units, supplied immediately from stock a ice level refers to the number of units Service Level tly on hand. For example, if annual 95 percent service level means that 950 can be ind 50 units are short. (This concept nrders are small and randomly distributed—one or several at a Vine, This model would not apply, for example, where the entire annual demand might be sold to ust s ew Socaas since we need enough data points to approximate a normal distribution The discussion in this section.on service levels is based on a statistical conce| known as Expected z or E(2).£(:) isthe expected number of prits short donne each jead time. Here, assume that the demand is normally distributed. To compute service level, we need to know how many units are short, For example, ome that the average weekly demand for an item is 100 units with a standard jation of 10 units. If we have 110 units at the beginning of a week, how many will we expect to be short? To do this, we need to summarize the probability that II are jemanded (1 short), the probability that 112 are demanded (2 short), plus the prob- ablity that 113 are demanded (3 short), and so on. This summary would give us the mber of units we would expect to be short by stocking 110 units. While the concept is simple, the equations are impractical to solve by hand. Fortu- ely, Robert Brown has provided tables of expected values (Exhibit 15.6). Exhibit 15.7 plots the numbers in Exhibit 15.6. This shows the expected number of sits short each order cycle (whether it is a periodic model P or an order quantity ‘node! Q). Using our previous example, suppose the average demand for an item was 0 units and the standard deviation’ of that demand was 10 units. In ne vnust multiply the vertical axis by 10 because the chart is based on a standard deviation daarnit: Reading from either the numbers in Exhibit 15.6 or the plot of these numbers in Exhibit 15.1, atz = 1, f we camry asafety stock of 10 units (one a deviation), we should expect to be out of stock just 83 units total. (.083 times a because the exhibits are based on a standard deviation of 1). Because normal demans j during the period is 100 and we were only short 3 (Iss than one nit), our service level is 100 — .83, or 99.17 pert ae Ifin the same example we did not c sve would be short 3.99 units (.399 times 10). Our service or 9 \inus one 1 . my ree ee same example, note that if we have a safety stock of mit gain, frot -xample, it the week rather * than 1007 ‘At90 units we would run short he beginning ofthe ¢ 89.17 percent. Carrying this further, if we have 80 units a Be ee ; 130 units: an ve will be short 20.08 units: if we have 70 units, Wt will be shor Because these exhibits are base’ don a stand: 10 do is to multiply the figures bY mand was 550 units and the aincees would give a .5 standard deviation * short of .198 X 36 = 7.128 umtt ; (550 — 7.128)/850 = 98:7 percent: afety stock (i.e, order just 100 units), level would be 100 ~ 3.99.ie Section Four ree 0399 Expected Number 0351 Onn of Stock versus =2.00 0.307 the Santer : “190 | 0267 Devin (Tne able ni 180 | 0230 040 fmm 110 Beinn a Be “hw | 016 00 a “150 | 0148 070 3700 ter 140 | 0120. 080 shoo tus 130 | 0100 090 300 te 120 | 0083 100 300 tw 110 | 0.069140 3300 ass 1.00 | 0086 1.20 ao 090 | 00% 1.30 tym 080, 0037 140 oss 0700029130 07-08 0028160 ome -050 0018 170 00 «9-020 ° 0014 1.80 os7 -030 (OO 1.90 oso -0.20° 0.008 2.00 asi 0.10 0.006 2.10 0399 = 0.000.005 2.20 jumber of standard deviations of safe!) stoex = Expected number of units shor: Source: Revised from Robert G. Brown. Decis.on Rules for Inventors Management (NEw York: Holt, Rinehart & Wissen. 19671. pp 25503 soa stedsummarize the preceding discuss) slsimply convent he sandal garasst fe service level approsch, gnit. Then, using Exhibit 15,6, m associated with the demand to we cae ticular service level In the caseot tren os sed the Standard normal distribution direae ty oF stocking out a istribution direet 18 OUL approach, we just of standard deviations of s; ly (Appendix D) to det st rhanage of henge ka cu ot deel hy Te aa ie nal turner of ca SL apbrOGch s that safety stock is determined based on he aca numberof Unis We desi lo deliver o our custome ae ions further wit : the fxed-order quantity mode and the Pee ene basic model types, ‘son eto how aeteey egmeweats ie able levels of customer service while minimizing inventory investment In our exant- ples, we will demonstrate the service lev Ry ee es approach to calculating safety sock: For tose who prefer to use the probaly ofsioeKoutappreach, commonly weeds vles afe 1.64 for 95 percent probability and 2.0 for 98 percent probability a base of one planned number of units short for a Fixed-Order Quantity Model with Specified Service Level A fixed-order quantity system perpetually monitors the inventory level and places a new order When stock reaches some level, R. The danger of stockout in this model rs only during the lead time, between the time an order is placed and the time it icreceived. As shown in Exhibit 15.8, an order is placed when the inventory position Jrops to the reorder point, R. During this lead time (L), a range of demands is possible. This range is determined either from an analysis of past demand data or from-an estimate (if past data are not available). : ‘The amount of safety stock depends On the service level desired, as previously discussed. The quantity to be ordered, Q, is calculated in the usual way considering the scinand, shortage cost, ordering cost, holding cost, and so forth. A fixed-order quan- tiy model can be ‘used to compute Q such as the simple Qj. model previously ‘lecusced. The reorder poigt is then set to cover the expected demand during the lead time plus a safety stock determined by the desired service evel Thus. the ke) differ- ten a fixed-order quantity model, where demand is known and one were Fae eencevtain is in computing the reorder point. The order quantity isthe fame oone see. The uncertainty element is taken into account inthe safety stock, The reorder point is ea ‘ (15.5) Fixed-Order Quantity Model 2 Number of units on handR d = Average daily . c. Lead time in days (time betweet standard deviations for a ion of usage during lead time effect is to place demand during the lead time. I and safety stock was computed to when 25 units remained. The greater the safety Computing , 7, and z Demand during the lead time to receive a tle eerie really an estimate or forecast of what is expected. It may be a single ffor example. if the lead time is a month, the demand may be taken as the year’s demand divided by 12). or it may be a summation of expected A lead time (such as the sum of daily demands over a 30-day lead time}. For the demand situation, d can be a forecasted demand using any of the models m= CI 13 on forecasting. For example, if a 30-day period was used to calculate d- simple average would be where n is the number of days. The standard deviation of the daily demand is. per day. If our lead time to get an order is five the five-day period. b ox deren = VOOF + (OF + (OF + (10 Next, we need to compute z.15 Inventory Systems for Independent Demand 597 comners eae ee level of P. (For example, P might be .95.) In the demand. Iwe ordered eal ei ct D ils © 0.05D, where D is the ana oa ‘nits each time, we would be placing D/Q orders per 5 ated heatey een 2e Mlistdtor’; Miy'R( shaventaege elie Gale - The expected number of units short » the expected number of units short is eons x Annual _ Number short ~~ Number of rt demand ~ per order orders per year Sheet i aae(c,. | x 2, which simplifies to 2 (1 ~ Pio EQ @ Ge (15.9) where P= Service level desired (such as satisfying 95 percent expressed as the fraction .95 of demand from items in stock) ‘raction of demand unsatisfied nnual demand tandard deviation of demand during lead time nomic order quantity calculated in the usual way (such as,Q = V2DS/H) E(;) = Expected number of units short each order cycle from a normalized table where o = 1 Note that D (annual demand) drops out of Equation 15.9. This is because E(z) is the number short each order cycle. (There are D/Q cycles per year.) ‘We now compare two examples. The difference between them is that in the first, the variation in demand is stated in terms of standard deviation over the entire lead time, while in the'second, it is stated in terms of standard deviation per day Quantity Consider an economic order 1,000 units, economic order quantity 195, the standard deviation of demand = 15 days. Determine the reorder mexample 15.3 Economic Order quantity case where annual demand D Q = 200 units, the desired service level P during lead time o, = 25 units, and lead time L point. solution In our example, 7 = 4 (1,000 over a 250-workday year), and lead time is 15 days. We use the equation R=dL+m = 4(15) + 2(25) ite i for E(z) and look this value up in the ¢ en 0 To ind wee 9 serie el? «95, and standard deviation ur problem : ; demand during lead time = 25- Therefore vn 1 je Oe BO ae A \ ca Seasdeciiem = Exhibit 15.6 and E(z) = 4, we find z = 0. Compete : From A ee R = (15) + 2(25) = 60 + 0(25) = 6Oumits ~~ s =e hand gets down to 60 units, order 200 n Tis eye that ae ne can calcalate the number of units a ee per yeat to see if it really is 95 percent. (2) is the expected are erat nseder based on a standard deviation of 1. The number short on ster for our problem is E (2), = 4(25) = 10. Because there are five orders per £1.000/200), this results in 50 units short. This yerifies our achievement of percent service level, because 950 out of 1,000 demand were filled from stock. Te 15.4 Order Quantity and Reorder Point Daily demand for ; Esule tes luct is normally heb ‘with a mean of 60 and standard deviation of 7. The source of supply is reliable and maintains a constant lead time of six days. The cost of placing the order is $10 and annual holding costs are $0.50 per unit. There no stockout costs, and unfilled orders are filled as soon as the order arrives. sales occur over the entire year. Find the order quantity and reorder point to 95 percent of the customers from stock on hand. solution Inthis problem we need to calculate the order quantity Q as well as the reorder point R. c : d= 60 S = $10 0% = 7. H = $0.50 D = 60365). = L=6 The optimal order quantity is _ PbS _ _ [2(60)365(10) _ 4 : Sone owaie 00 = V876,000 = 936 units To compute the reorder point, we need to calculate the amount of product us during the lead time and add this to the safety stock, « : The standard deviation of demand during the lead time of six days is calculatet from the variance of the individual days. Because each day’s demand is indepen z = 4/04, = Vey = 172 3 i Se & Next we need to know how many standard deviations are needed for a service level. As previously defined, Ql = P) e EQ). = = on Therefore 936(1 — .95) 1S eee Type 2.721 a From Exhibit 156, imerpolating at E(2) = 2.721, 2 = 2.72 E() =Out to be negative. This means that When our inventory position dropped = 360), we would have had a higher oe = 501365) Sage this example by noting that we would place 23.4 TD TTD pee )/ 936]: Each petiod would experience 46,8 wats ofr afenrce sce evel deem Me would be out of stock 1,095 unite per year (46,8 23-4), ce level. therefore, is 0.95 as we intended [(21.900 — 1 095)/21,900}. = oe :bon in these two examples, this technique of determining safety tack levels atively — Straightforward. It allows us to control inventory to meet our d service levels. ' @ FIXNED-TIME PERIOD MODELS fixed-time period system, inventory is counted only at particular times, such as ery week or every month. Counting inventory and placing orders on aperiodic basis irable in situations such as when vendors make routine visits to customers and e orders for their complete line of products, or when buyers want to combine orders save transportation costs. Other firms operate on a fixed-time period to facilitate nning their inventory count; for example, Distributor X calls every two weeks and oyees know that all Distributor X’s product must be counted xed-time period models generate order quantities that vary from period to pe- nod, depending om the usage rates. These generally require a higher level of safety k than a fixed-order quantity system. The fixed-order quantity system assumes, continual counting of inventory on hand, with an order immediately placed when the reorder point is reached. In contrast, the standard fixed-time period models assume that inventory is counted only at the time specified for review. It is possible that some large demand will draw the stock down to zero right after an order is placed. This condition could go unnoticed until the next review period. Then the new order, when shout the laced. 1. Thus. it is possible to be out of stock throug! seen ae ees me ar onder led time, L, Safety stock, therefore, must protect view period, T. and of ws agri eae pie the review period itself as well as during the lead time from order placement to order receipt. Fixed-Time Period Model with Specified Service Level In a fixed-time period system, reorders are placed at the time of review (P), and the safety stock that must be reordered is i Safety stock = 207+ Exhibit 15.9 shows a fixed-time period constant lead time of L. In this casesSection Four Managing the Supply Chain on tand_| Place (in units) Safety ai ee Time t ‘The quantity to order, q, is Average demand Inventory currently 4 Order _ over the vulner- , Safety _ on hand (pluson (15.19) quantity able period stock order, if any) TS qe Shar Da zor IT where q = Quantity to be ordered T = The number of days between reviews ead time in days (time between placing an order and receiving it) d = Forecasted average daily demand lumber of standard deviations for a specified service level G;. = Standard deviation of demand over the review and Jead time 7 = Current inventory level (includes items on order) Note: The demand, lead time, review period, and so forth can be any time units such as days, weeks, or years so long as it is consistent throughout the equation. Inthis model, demand (d ) can be forecast and revised each review period if desired, or the yearly average may be used if appropriate. We assume that demand is no distributed. ; The value of z can be obtained by solving the following eqi reading the corresponding z value from Exhibit 15.6: _ aT = P) Oren E where E(2) = Expected number units short from a normalized table where _P = Service level desired expressed as a fraction (¢.g., 95 percent as 9 aT = Demand during the review period where d is daily demand and Tis number of days a + = Standard deviation over the review period and lead time beagle 135 Quantity 10 Order Daily demand fora pro ee E aa 'd deviation of three units. The review period is 30 days, is 14 days. Management has set a policy of satisfying 98 percent of d in stock. At the beginning of this review period, there are 150 units How many units should be ordered? 4 so= — ‘Chapter 15 Inventory Systems for Independent Demand : solution The quantity to order is G2 dT +1) + co. -1 = 100 + 14) + 20,44 — 150 Before we can complete the solution, we we need to find oy, and z, To find use the notion, as before, that the standard deviation of setuenoe of saad ° Ores ow, (15.12) Because each day is independent and ay is constant rn = VT + Deh = VG0 > GF = 19.90 Now to find =. we first need to find (2) and look this value up in the table. In this demand during the review period is dT, so dT(1 ~ P)_ 10030)(1 — 98) ook 19.90 rom Exhibit 15.6 at E(z) = 0.302, by interpolation. = = 24 The quantity to order, then, is = d(T + L) + z07,, — E() = 0.302 10(30 + 14) + .21(19.90) — 150 = 294 units To satisfy 98 percent of the demand for units, order 294 at this review period. m i SPECIAL PURPOSE MODELS The fixed-order quantity and the fixed-time period models presented thus far differed » their assumptions but had two characteristics in common: (1) The cost of units rema any order size; and (2) the reordering process was continu- ree nua yn eae) ose 4 continue Thes section presents two new mi uantity when unit price changes with ‘sometimes called a static model) in-w off each time. This type of model is amen: vodels. The first illustrates the effect on order order size. The seconds a single-period model which ordering and stocking require a cost able to solution by marginal analysis. ‘ather than a per-unit change. For example. "99 screws $1.60 per 100, and $13.50 pe 1,000 To St any tem to order, we simply solve fot nd at the point of pri¢e change. But te 2 bY the formula are feasi 4) 4s that the optimal « “Ould be impossible,ase a a e Managing te Surly ale ee Section Four order Quanity Models na Three Price Break S 602 +s for Three Separate Ord (RONMENT ne esse) eee Ondering cost and holding ie 300 300400. S00 GO. 700-800. 900 1,000 1,100 1,200 Units Price-Brea! ‘The total cost for each feasible economic order quantity and price-bres Quantity oad quantity is tabulated, and the Q that leads to the minimum cost is the opti size. If holding cost is based on a percentage of unit price, it may not be compute economic order quantities at each price. Procedurally, the large ‘quantity (lowest unit price) is solved first; if the resulting Q is valid, thatis the: Ifnot, the next largest quantity (second lowest price) is derived. If thatis f cost of this Q is compared to the cost of using the order quantity at the above, and the lowest cost determines the optimal Q. Looking at Exhibit 15.10, we see that order quantities are solved from rij or from the lowest unit price to the highest, until a valid Q is obtained, Ther quantity at each price break above this Q is used to find which order least cost—the computed Q, or the Q at one of the price breaks. m example 15.6 Price-Break Consider the following case, D = 10,000 units (annual demand) 5 = $20 to place each order i = 20 percent of cost (annual carrying cost, storage, interest, obsole ‘C= Cost per unit (according to the order size; orders of 0 to 499 Per unit; 500 to 999, $4.50 per unit; 1,000 and up, $3.90, What quantity should be ordered?”. Chapter 15 Inventory Systems for ZB B= Om cae om whee mn ets. ong where Break ‘i C= $450 oie Helin Se £ es Relevane Costs in 2 ( Three Price Break #) Sic 666 - 7 4,000 Model G 1003.5 0 s390 Ordering cost $299.70 p = $390 D ) a (as Na fenble 1000030) fasouan a 3 Not feasthe “OOS nag os ome = tem cost (DC). 00 ne 10.00044,50) ee $45,599.70 asta _—_———oooo Solving for the economic order size, we obtain @ C = $3.90,Q = 716 Not feasible 2 C = $4.50, = 666 Feasible, Cost = $45.599.70 Check Q = 1,000 Cost = $39,590 Optimal solution In Exhibit 15.10, which displays the cost relationship and order quantity range. note { most of the order quantity-cost relationships lie outside the feasible range and ‘nly a single, continuous range results. This should be readily apparent because. ple, the first order quantity specifies buying 633 units at $5.00 per unit. However. if 633 units are ordered, the price is $4.50. not $5.00. The same holds true or the third order quantity, which specifies an order of 716 units at $3.90 each. This $3.90 price is not available on orders of less than 1.000 units. Exhibit 15.1] itemizes the total costs at the economic order quantities and at the price breaks. The optimal order quantity is shown to be 1,000 units. One practical conisideration in price-break problems 1s that the price reduction from volume purchases frequently makes it seemingly economical to order amounts larger than the Qjyy. Thus, when applying the model. we must be pafticularly careful to obtain a valid estimate of product obsolescence and warehousing costs ‘Some inventory situations involve placing orders to cover Single-Period Models ad items at frequent internals. Sometimes iod or to cover short-h cae re Ee ee problems (for example, how many papers should Jution through the classic eco- fh day), they are amenable to Sol 5 ney onde gna anys The optimal stocking Som 0 TH i i unit are analysis, occurs at the point where she benefits derived from carrying the Ce Pies less than the costs for that unit. Of course, specific benefi he selection of the le, we mm: > costs depends on the problem. For exam ©. ; i ous marginal 10sS_ ens Rong css or as we vel ORE) MATE abs When stocked items ete he Sale oF use ofthe last unit egal t i re é terms. sock ha gui We gems symm MSS condition where MP = ML. where i ési is is also valid when we are dealing with probabil Ne ae we are looking at expected profits and e3 renetjucing probabilities, the marginal profit-marginal loss equation b P(MP) = (1 ~ P)ML i 1s bel id 1 ~ P is the pro obability of the unit’s being sold an Pi as because one or the other must occur, (The unit is sold or Then, solving for P, we obtain ML P= MP + ML . This equation states that we should continue to increase the size of the so long as the probability of selling the last unit added is equal to or greater ratio ML/(MP + ML). ‘ Salvage value, or any other benefits derived from unsold goods, can included in the problem. This simply reduces the marginal loss, as the fo example shows. $ mexample 15.7 Salvage Value A product is priced to sell at $100 per and its cost is constant at $70 per unit. Each unsold unit has a salvage value of $2( Demand is expected to range between 35 and 40 units for the period; 35 u definitely can be sold and no units over 40 will be sold. The demand probabilities an the associated cumulative probability distribution (P) for this situation are sho Exhibit 15.12. 3 The marginal profit if a unit is sold is the selling price less the cost,-or $100 — 70 = $30. The marginal loss incurred if the unit is not sold is the cost of the unit salvage value, or ML = $70 — $20 = $50. i How many units should be ordered? a solution The optimal probability of the last unit being sold is - iM SncaliteSU MP +ML 30450 According to the cumulative probability table (the 1a8t column in Exkjbit 1 the probability of selling the unit must be equal to or greater than 0.625, 0 3 should be stocked. The probability of selling the 37th unit is 0.75, The from stocking the 37th unit is the expected marginal profit minus the ex marginal loss. : , a P = 0.625 ‘ Net = P(MP) — (1 — P)(ML). ee = 0.75($100 — $70) — (1 ~ 0.75)($70 = $2¢ = $22.50 — $12.50 = $10 hit ce Our gdChapter 15 Inventory Systems for eee ® ‘of Units Probabilit wy y of Demanded ‘The Dour The Vat Pebatiny ss Sallng O10 1035 % ous Pt naa 4 hs n Fe : 025 a = 40 om ‘9 025 a oe 4 0.10 9 41 of more ° 605 ‘Demand and Cumulative Probabilities ” au) a) Net Prat o o Probability Expected Marginal Expected Marginal ofthe ath Ualts of Prsbabilty ot Seling”—"Prottatatntnt,—Tesstamn Ga Unit Demand of Demand ath Unit ‘P1000 — 70) (= P70 = 20) (MP) ~ (ML) 5 O10 100 30 so $30.00 6 ous 090 2 s 2 " 2s ons 250 1250 0 18 02s 050 is » 10 9 ous as 2.50 x70 =0 ‘0 oo 0 : 6 2 41 0 0 —————— eee Note’ Expected marginal profit i the selling price of $100 less the unit cost of $70 times the probability the Fapected marginal los is the unit cost of $70 less the salvage value of $20 times the probability the unit will no @ MISCELLANEOUS SYSTEMS AND I ES. and shortage costs is difficul—sometumes es. unrealistic. For example, Ex- med linear to the real case where Obtaining actual order, sep, carrying, and shor impossible, Even the assumptions are sometit hibit 15.14 compares ordering costs that are assumed “tet | every addition of a staff person emis FE aod problems; maintaining adequate All inventory systems are plagued 9} it zcords of stock on hand col nvemony se em and ennring that accurate ene O68 tre kept. In this section, we present three simple SGT cy, ABC anal- eps pe an ing inventor ysis (a method for analyzing inven! ; ). nique for improving inventory record accuracy wont will be sald be sold ABC Analysisiu aes Cost to Place Orders versus the Number of ‘Orders Placed: Linear ‘costs and Normal Reality System System Ordering Number of orders ‘Number of orders this is a P model. For example, the maximum inventory level (which we will e can be computed based on demand, ordering costs, and shortage costs, B takes time and costs money to place an order, a minimum order of size established. Then, whenever this item is reviewed, the inventory position (we it) is subtracted from the replenishment level (M), If that number (cal it q) is to or greater than Q, order q. Otherwise, forget it until the next review period, formally, 5 q=M-—# If q = Q, order g. Otherwise, do not order any. Two-Bin System In a two-bin system, items are used from one bin, and second bin provides an amount large enough to, ensure that the stock can be ished. In Exhibit 15.1, this is a Q model. Ideally, the second bin would contain 3 amount equal to the reorder point (R) calculated earlier. As soon as the second bi supply is brought to the first bin, an order is placed to replenish the second | Actually, these bins can be located together. In fact, there could be just one bin wit a divider between. The key to a two-bin operation is to separate the inventory so} Part of it is held in geserve until the rest is used first. One-Bin System A one-bin inventory system involves periodic replenis! matter how few are needed, At fixed periods (such as weekly), the inventory is b up to its predetermined maximum leyel. The one bin is always replenished, therefore differs from the optional replenishment system, which only reorders the inventory used is greater than some minimum amount. This is a P model hibit 15.1, ABC Inventory Planning Maintaining inventory through counting, placing orders, receiving stock, takes personnel time and costs money. When there are limits on these logical move is to try to use the available resources to ct inventory i In other words, focus on the most important items in inventory. In the nineteenth century, Villefredo Pareto, in a study of the d in Milan, found that 20 percent of the people controlled 80 per logic of the few having the greatest importance and the man has been broadened to include many situations and is terChapter 15 Inventory Systems for Independent Demand # tem Nu exhibit (3.1 | 22 . 4 $95,000 68 75,000 Ate ve a j ny . vnnual Usage a 25,000 10.7 ented aks % 15,000 6s % 13,000 56 ‘ 4 7.300 32 6 1.500 06 a 800 7 egy 23 42: > 2 5 02 on ‘most of our decisions are relatively unimportant, but tainly true in inventory ventory systems (where a few count for the bulk of our investment). ie ene Any inventory system must specify when an order is to be placed for an item and how many units to order. Most inventory control situations involve so many items that {us not practical to model and give thorough treatment to each item. To get around his problem. the ABC classification scheme divides inventory items into three group- gs: high dollar volume (A), moderate dollar volume (B). and low dollar volume (C). Dollar volume is 4 measure of importance; an item low in cost but high in volume can be more important than a high-cost item with low volume. ABC Classification If the annual usage of items in inventory is listed according to dollar volume, generally, the list shows that a small number of items account for a large dollar volume and that a large number of items account for a small dollar volume. Exhibit 15.15 illustrates the relationship. The ABC approach divides this list into three groupings by value: A items consti- tute roughly the top 15 percent of the items, B items the next 35 percent. and C items the last 50 percent. From observation, it appears that the list in Exhibit 15.15 may be meaningfully grouped with A including 20 percent (2 of the 10), B iota She cent, and C including 50 percent. These points show clear eae nae sections. The result of this segmentation is shown in Exhibit 15.16 and plotted i Exhibit 15.17. Segmentation may not always occur so neatly. The seer ak ul eS Separate the important from the unimportant. Where pe ties ene a yee cat on the particular inventory under question and on how rc Peet, able. (With more time, a firm-could define larger A ono ees The purpose of classifying items into groups is 10 ¢5 en ee ae of ontsol over each item. On a periodic basis for example ener end C clearly controlled with weekly ordering. B items ee es ee ae items may be ordered monthly or bimonthly. Note 4 Telated to their classification. An oe Combination of either low cost and high Citems may have a low dollar volume n automobile service station, gasoline ‘plenishment; tires, batteries, oll. Bre#tt Ordered every two to fourABC Inventory Cassifiaton iowentory valve fOr tach group versus the group's portion of the oa bs) -y Accuracy Managing the Supply Chain ‘section Four tem Number Annual Dollar Usage 27, 03, 82 $4, 36,19, 23, 41 B items 20 40 Percent of total list of different stock items wiper blades, radiator caps, hoses, fan belts, oil and gas additives, car Wax, am forth. C items may be ordered every two or three months or even be allowed to out before reordering because the penalty for stockout is not serious. 4 ‘Sometimes, an item may be critical to a system if its absence creates a sizable lo In this case. regardless of the item’s classification, sufficiently large stocks sho Kept on hand to prevent runout. One way to ensure closer control is to designate item an A or a B. forcing it into the category even if its dollar volume does not warra such inclusion Inventory Accuracy and Cycle Counting : Inventory records usually differ from the actual physical count; inventory ac refers to how well the two agree. Companies such as Wal-Mart (see Breakthr Box) understand the importance of inventory accuracy and expend considerable ensuring it. The question is, How much error is acceptable? If the record balance of 683 of part X and an actual count shows 652, is this within Suppose the actual cotint shows 750, an excess of 67 over the record; is this better? § Every production system must have agreement, within some specified ra tween what the record says is in inventory and what actually s in inventory. many reasons why records and inventory may not agree. For exagpt stockroom area allows items to be removed for both legitimate and purposes. The Iegitimate removal may have been done in a hurry and recorded. Sometimes parts are misplaced, turning up months later. stored in several locations, but records may be lost or the location, rectly. Sometimes stock replenishment orders are recorded as f they never were. Occasionally, a group of parts is recorded as rel tory, but the customer order is canceled and the parts are replaced i canceling the record. To keep the production system flowingChapter 15 Inventory Systems Torre a Grand Seale | .d, the physical count h the book inventory, | internal audit | Inventory on [pes at Wal-Mart is precision on a | ter the inventory Is complete Sargantuan scale, like a maneuver of the team reconciles its findings wit \ulantic Fleetora dam project inthe Yangtze River. The results are reviewed later by the # 7 i; bs fuling on Wal-Mart's Tax Court case.» department, " ludge David Laro provided a behind.the-scenes Inventories are taken every 11 to 13 months, and , ok at the operation. Its effectiveness, he says, has most occur from March through September: inventory’ | led “many other companies, both domestic and is never taken in November or December. when it foreign” “10. seek Wal-Mart's advice on would interfere with the Christmas season. or in the natn pe acca ecary nen eroployeesare rene on ee takes four, to six weeks and busy with exchanges and returns ays in advance, the chain's internal avdit The job can be made no casier by the fact that =_—_ML eek MP + ML (5.14) SOLVED PROBLEMS SOLVED PROBLENE ) hems purchased from a vendor rdor cost $30 1,000 units. If it cost gost $70 each, and the for ; see il peace SS Pee toca a ea cheat years dem Pe uni er Yar. what quant sn for more units and the storage cost a. What is the total ordering cost fore oroe7°d each time? What isthe foal storage eoa ter yee? Solution The quantity to be ordered each time is of 2DS _— /2(1.000)5 Va Sayre: or a. The total ordering cost for a year is D 1.00 ° 30 ) = $100 b. The storage cost for a year is 0 Sy = 2is4) = s100 SOLVED PROBLEM 2 : Daily demand for a product is 120 units. with a standard deviation of 30 units. The review 7 k, period is 14 days and the lead times 7 days. At the time of review there are 130 units in stoc 199 percent of all demand is to be satisfied from items in stock. how many units should be ordered? Solution ; Or. = Vda + TBO = V18,900 = 137.5 _ 120(14)(1 = 99) 9 439 137.5 : (hen From Exhibit 15.6. = = 80 | : a(T + L) + 0p = 12014 + 7) + Biel ii E(2) q — ce _‘Section Four manager — in late this afternoon when thei come m Paina |. How many units should be o nT ea pacers! = 180, 7 = 14,L = 7, = 20 = VIG = 23 Or = = EQ = From Exhibit 15.6, 2 = 12a at) — 180 258.4 units q REVIEW AND DISCUSSION QUESTIONS fv tween dependent and independent demand in a McDonald's, in i pee re of personal copies, and in a pharmaceutical supply Distinguish between in-process inventory, safety stock inventory, and seasonal inventory. : Discuss the nature of the costs that affect inventory size. Under which conditions would a plant manager elect to use a fixed-order quan model as opposed to a fixed-time period model? What are the disadvantages of us fixed-time period ordering system? 5. Discuss the general procedure for determining the order quantity when price b are involved. Would there be any differences in procedure if-holding cost were a percentage of price rather than a constant amount? 5 6. What two basic questions must be answered by an inventory-control decision . Discuss the assumptions that are inherent in production setup cost, ordering: carrying costs. How valid are they? 5 8. “The nice thing about inventory models is that you can pul it so long as your cost estimates are accurate.” Comment. 9. Which type of inventory system would you use in the following situations? 4. Supplying your kitchen with fresh food. aa 5. Obtaining a daily newspaper. Buying gas for your car. To which of these items do you impute the highest stockout cost? ‘Why is it desirable to classify items into groups, as-the ABC classification What kind of policy or procedure would you recommend to improve the ‘operation in a department store? What advantages and disadvantages do have a vis-a-vis the department store inventory operation described in ry 3 4 ill one off the shelf 10, 1. PROBLEMS