第六單元:Inventory Management: Safety Inventory ( I ) Inventory Management: Safety Inventory ( I ) 郭瑞祥教授 【本著作除另有註明外,採取創用CC「姓名標示 -非商業性-相同方式分享」台灣3.0版授權釋出】 1 Safety Inventory ► Safety Inventory is inventory carried for the purpose of satisfying demand that exceeds the amount forecasted for a given period. ► Purposes of holding safety inventory –Demand uncertainty –Supply uncertainty Inventory Average Inventory Cycle Inventory Safety Inventory Time 2 Planning Safety Inventory ► Appropriate level of safety inventory is determined by 》 Uncertainty of both demand and supply – Uncertainty increases, then safety inventory increases. 》 Desired level of product availability Desired level of product availability – increases, then safety inventory increases. ► Actions to improve product availability while reducing safety inventory 3 Measuring Demand Uncertainty ► Uncertainty within lead time – Assume that demand for each period i, i=1,….,k is normally distributed with a mean Di and standard deviation si . – The total demand during k period is normally distributed with a mean of P and a standard deviation of W : W= k P= 2Di s i=1 i +2 k i=1 Cov(i,j) = i>j k si2 +2 rsisj i=1 i>j –If demand in each period is independent and normally distributed with a mean of D and a standard deviation of sD , then P=KD W = k sD Coefficient of variation CV= s/m 4 k si2 +2 rsisj i=1 k sD i>j Measuring Product Availability ► Product fill rate ( fr ) – – CoolCLIPS網站 The fraction of product demand that is satisfied from product in inventory It is equivalent to the probability that product demand is supplied from available inventory ► Order fill rate – The fraction of orders that are filled from available inventory – Order fill rates tend to be lower than product fill rates because all products must be in stock for an order to be filled ► Cycle service level (CSL) – – – 5 The fraction of replenishment cycles that end with all the customer demand being met The CSL is equal to the probability of not having a stockout in a replenishment cycle A CSL of 60 percent will typically result in a fill rate higher than 60% Measuring Product Availability -- Page 5 ► Product fill rate ( fr ) – An order for a total of 100 palms and has 90 in inventory → fill rate of 90% ► Order fill rate – Customer may order a palm along with a calculator. The order is filled only if both products are available. ► Cycle service level (CSL) Order received – Don't run out of inventory in 6 out of 10 replenishment cycles → CSL = 60% – In the 40% of the cycles where a stockout does occur, most of the customer demand is satisfied from inventory On-hand inventory 0 → fill rate > 60% 6 Filled demand Cycle Microsoft。 Microsoft。 Unfilled demand Replenishment Policies ► A replenishment policy consists of decisions regarding – When to reorder – How much to reorder. ► Continuous review Q – Inventory is continuously tracked and an order for a lot size Q is placed when the inventory declines to the reorder point (ROP). ► Periodic review P – Inventory status is checked at regular periodic intervals and an order is placed to raise the inventory level to a specified threshold, i.e. order up to level (OUL) . 7 Replenishment Policies ► A replenishment policy consists of decisions regarding – When to reorder – How much to reorder. ► Continuous review Q – Inventory is continuously tracked and an order for a lot size Q is placed when the inventory declines to the reorder point (ROP). ► Periodic review P – Inventory status is checked at regular periodic intervals and an order is placed to raise the inventory level to a specified threshold. 8 Continuous Review System ► The remaining quantity of an item is reviewed each time a withdrawal is ► made from inventory, to determine whether it is time to reorder. ► Other names are: Reorder point system, fixed order quantity system ► Inventory position IP = OH+SR-BO 》IP = inventory position 》OH = on-hand inventory 》SR = scheduled receipts (open orders) 》BO = units backordered or allocated ► Decision rule – Whenever a withdrawal brings IP down to the reorder point (ROP), place an order for Q (fixed) units. 9 Continuous Review System ROP = average demand during lead time + safety stock IP On-hand inventory IP Order received Order received Q OH OH ROP Order placed Order placed L1 TBO1 10 L3 L2 TBO2 TBO3 Time Continuous Review System ROP = average demand during lead time + safety stock IP On-hand inventory IP Order received Order received Q OH OH ROP Order placed Order placed L1 TBO1 11 FIX TBO2 L3 L2 TBO3 Time Continuous Review System ROP = average demand during lead time + safety stock IP On-hand inventory IP Order received Order received Q OH OH ROP Order placed Order placed L1 TBO1 12 L3 L2 TBO2 TBO3 Time Example Given the following data Average demand per week, D = 2,500 Standard deviation of weekly demand, sD =500 Average lead time for replacement, L = 2 weeks Reorder point, ROP = 6,000 Average lot size, Q = 10,000 ► Safety inventory,ss =ROP-DL=6,000-5,000=1,000 ► Cycle inventory =Q/2=10,000/2=5,000 ► Average inventory =5,000+1,000=6,000 ► Average flow time = Average inventory / Throughput=6,000/2,500 =2.4weeks 13 Evaluating Cycle Service Level and Safety Inventory CSL= Prob (Demand during lead time of L weeks ROP) Demand during lead time is normally distributed with a mean of DL and a standard deviation of sL DL = DL and s L = Ls D ► CSL=Function ( ROP,DL,sL ) ROP=DL+Z LsD ss=z LsD z=Fs-1(CSL) 14 CSL Finding Safety Stock with a Normal Probability Distribution for an 85 Percent CSL ? CSL = 85% 1 Average demand during lead time Probability of stockout (1.0 - 0.85= 0.15) 4:->ROP ROP 3 2 z sL Safety stock = z sL 15 Evaluating Cycle Service Level and Safety Inventory CSL= Prob (Demand during lead time of L weeks ROP) Demand during lead time is normally distributed with a mean of DL and a standard deviation of sL DL = DL and s L = Ls D CSL=Function ( ROP,DL,sL ) ROP=DL+Z LsD ss=z LsD z=Fs-1(CSL) 16 Example Given the following data Q = 10,000 ROP = 6,000 L = 2 weeks D=2,500/week, σD=500 ► DL=DL= 2x2,500=5,000 ► sL= L sD = 2 x500=707 ► CSL=Proability of not stocking out in a cycle =F(ROP, DL, sL )=F(6000,5000,707) =NORMDIST(6000,5000,707,1)=0.92 17 Normal Distribution in Excel Commands (Page 12) ► F ( x, m,s ) = NORMDIST ( x, m,s ,1) f ( x, m,s ) = NORMDIST ( x, m,s ,0) F 1( p, m,s ) = NORMINV ( p, m,s ) ► Standard Normal Fs ( x ) = NORMDIST ( x ) or NORMDIST ( x,0,1,1) fs ( x ) = NORMDIST ( x,0,1,0) Fs1( p ) = NORMINV ( p ) 18 Normal Distribution in Excel (Demo) 19 臺灣大學 郭瑞祥老師 Example Given the following data Q = 10,000 ROP = 6,000 L = 2 weeks D=2,500/week, sD=500 CSL=0.9 ► DL=DL= 2x2,500=5,000 ► sL= L sD = 2 x500=707 ► ss=Fs-1(CSL) =F(ROP, DL, sL )=F(6000,5000,707) =NORMDIST(6000,5000,707,1)=0.92 20 Example Given the following data D=2,500/week sD=500 L = 2 weeks Q = 10,000, CSL=0.9 ► DL=DL= 2x2,500=5,000 ► sL= L sD = 2 x500=707 ► ss=Fs-1(CSL)xsL=NORMDIST(CSL)xsL =1.282x707=906 ► ROP= 2x2,500+906=5,906 21 Example Given the following data D=2,500/week sD=500 L = 2 weeks Q = 10,000, CSL=0.9 DL=DL= 2x2,500=5,000 sL= L sD = 2 x500=707 ss=Fs-1(CSL)xsL=NORMDIST(CSL)xsL =1.282x707=906 ROP= 2x2,500+906=5,906 22 臺灣大學 郭瑞祥老師 Periodic Review System ► Other names are: fixed interval reorder system or periodic reorder system. ► Decision Rule Review the item’s inventory position IP every T time periods. Place an order equal to (OUL-IP) where OUL is the target inventory, that is, the desired IP just after placing a new order. ► The periodic review system has two parameters: T and OUL. ► Here Q varies, and time between orders (TBO) is fixed. 23 Periodic Review System OUL OUL IP On-hand inventory Order placed IP Order received Order placed Q1 Q3 Q2 IP1 OH OH IP3 Order placed IP2 L T Protection interval 24 L L T Time Finding OUL ► The new order must be large enough to make the inventory position, IP, last not only beyond the next review, which is T periods from now, but also for one lead time (L) after the next review. IP must be enough to cover demand over a protection interval of T + L. ► OUL = = 25 Average demand + during protection interval (T L)D Safety stock for protection interval Fs1 (CSL) T L sD Selecting the Reorder Interval (T ) ► Administratively convenient (such as each Friday) ► Approximation of EOQ EOQ T= (52) weeks D ► Example: Suppose D = 1200 /year and EOQ = 100 100 T= (52) = 4.3 weeks or 4 weeks 1200 26 Example Given the following data D=2,500/week sD=500 L = 2 weeks T= 4weeks CSL=0.9 ► DT+L=(T+L)D= (4+2)x2,500=15,000 ► DT+L= T+L sD= (4+2) x500=1,225 ► ss=Fs-1(CSL)xsT+L=Fs-1(0.9)xsT+L =1,570 ► OUL=DT+L+ss = 1,5000+1,570=16,570 27 Periodic System versus Continuous System Feature Order quantity When to place order Recordkeeping Q-constant When quantity on hand drops to the reorder level Each time a withdrawal or addition is made Q-variable When the review period arrives Counted only at review period Size of inventory Less than periodic system Larger than continuous system Factors driving safety inventory Demand uncertainty Replenishment lead time Type of items 28 Continuous review Periodic review system system Higher-priced, critical, or important items Demand uncertainty Replenishment lead time Reorder interval Evaluating Fill Rate Given a Replenishment Policy ► For a continuous review policy Expected shortage per replenishment cycle (ESC) is the average units of demand that are not satisfied from inventory in stock per cycle ESC = (X-ROP) f(x)dx X=ROP f (x) is density function of demand distribution during the lead time fr=1- ESC = Q-ESC Q Q ► In the case of normal distribution, we have 29 Evaluating Fill Rate Given a Replenishment Policy ► For a continuous review policy Expected shortage per replenishment cycle (ESC) is the average units of demand that are not satisfied from inventory in stock per cycle ESC = (X-ROP) f(x)dx X=ROP f (x) is the density function of demand distribution during the lead time fr=1- ESC = Q-ESC Q Q ► In the case of normal distribution, we have 30 Evaluating Fill Rate Given a Replenishment Policy ► For a continuous review policy Expected shortage per replenishment cycle (ESC) is the average units of demand that are not satisfied from inventory in stock per cycle ESC = (X-ROP) f(x)dx X=ROP f (x) is density function of demand distribution during the lead time fr=1- ESC = Q-ESC Q Q ► In the case of normal distribution, we have 31 Evaluating Fill Rate Given a Replenishment Policy ► For a continuous review policy Expected shortage per replenishment cycle (ESC) is the average units of demand that are not satisfied from inventory in stock per cycle ESC = (X-ROP) f(x)dx X=ROP f (x) is density function of demand distribution during the lead time fr=1- ESC = Q-ESC Q Q ► In the case of normal distribution, we have ss ss ESC = ss 1 Fs s Lfs s L sL = ss1 NORMDIST ss / s L , 0 , 1 , 1 s LNORMDIST ss / s L , 0 , 1 , 0 32 Proof ( x D )2 2s 2 1 L dx ESC = ( x ROP )f ( x )dx = ( x DL ss ) e L x =ROP x =DL ss 2 s L WIKIPEDIA 33 WIKIPEDIA Proof ( x D )2 2s 2 1 L dx ESC = ( x ROP )f ( x )dx = ( x DL ss ) e L x =ROP x =DL ss 2 s L Substituting Z=(X-DL)/sL and dx=sLdz , we have ESC = s z =ss / 34 L 1 z2 / 2 ( zs L ss ) e dz 2 sLdz Proof ( x D )2 2s 2 1 L dx ESC = ( x ROP )f ( x )dx = ( x DL ss ) e L x =ROP x =DL ss 2 s L 2 /2 1 L)/s Substituting Z=(X-D z 2 / 2L and dx=sLdz 1, we zhave e dz = ss e dz s L z z =ss / s 2 2 z =ss / s L L 1 z2 / 2 ESC = ( zs L ss ) e dz 2 z =ss / s L 35 Proof ESC = s z =ss / = ss 36 L 1 z2 / 2 ( zs L ss ) e dz 2 z =ss / s L 1 2 e z2 / 2 1 dz s L z z =ss / s 2 L e z2 / 2 dz Proof ESC = s z =ss / = ss L 1 z2 / 2 ( zs L ss ) e dz 2 z =ss / s L 1 2 e z2 / 2 = ss[1 Fs (ss / s L )] 37 1 dz s L z z =ss / s 2 L e z2 / 2 dz Proof ESC = s z =ss / = ss L 1 z2 / 2 ( zs L ss ) e dz 2 z =ss / s L 1 2 e z2 / 2 1 dz s L z z =ss / s 2 L = ss[1 Fs (ss / s L )] s L 1 w e dw 2 w =ss2 / 2 s 2L 1 = ss[1 Fs (ss / s L )] s L [ 2 38 e 1 ss 2 s L e z2 / 2 2 ] dz (note : w = z 2 / 2) dw=2zdz/2 dw=zdz Proof ESCderivation 1 0 z2 / 2 1(ss / s )2 ESC = ( zs L ss ) e dz L 1 1 2 1 2 z =ss / s = e e d = e L 2 2 ss / 2s L 2 2 =ss2 / 2s L2 2 1 1 z2 / 2 z2 / 2 e dz = ss e dz s L z z =ss / s 2 z =ss / s L L 2 = ss[1 Fs (ss / s L )] s L w =ss2 / 2s L2 1 = ss[1 Fs (ss / s L )] s L [ 2 39 1 2 e 1 ss 2 s L 2 ] e w 2 (note : w = z / 2) dw Proof ESC = s z =ss / = ss L 1 z2 / 2 ( zs L ss ) e dz 2 z =ss / s L 1 2 e z2 / 2 1 dz s L z z =ss / s 2 L = ss[1 Fs (ss / s L )] s L 1 w e dw 2 w =ss2 / 2 s 2L = ss[1 Fs (ss / s L )] s L [ 1 2 e 1 ss 2 s L 2 = ss[1 Fs (ss / s L )] s Lfs (ss / s L ) 40 e z2 / 2 ] dz (note : w = z 2 / 2) Evaluating Fill Rate Given a Replenishment Policy ► For a continuous review policy Expected shortage per replenishment cycle (ESC) is the average units of demand that are not satisfied from inventory in stock per cycle ESC = (X-ROP) f(x)dx X=ROP f (x) is density function of demand distribution during the lead time fr=1- ESC = Q-ESC Q Q ► In the case of normal distribution, we have ss ss ESC = ss 1 Fs s Lfs s L sL = ss1 NORMDIST ss / s L , 0 , 1 , 1 s LNORMDIST ss / s L , 0 , 1 , 0 41 Example For a continuous review system with the following data Lot size ,Q=10,000 DL=5,000 sL = 707 ► ss=ROP-DL=6,000-5,000=1,000 ► ESC= -1,000[1-NORMDIST(1000/707,0,1,1)] +707xNORMDIST(1000/707,0,1,1) =25 10,000-25 fr= 10,000 42 =0.9975 Excel-Demo For a continuous review system with the following data Lot size ,Q=10,000 DL=5,000 sL = 707 ss=ROP-DL=6,000-5,000=1,000 ESC= -1,000[1-NORMDIST(1000/707,0,1,1)] +707xNORMDIST(1000/707,0,1,1) =25 10,000-25 fr= 10,000 43 =0.9975 臺灣大學 郭瑞祥老師 Factors Affecting Fill Rate ► Safety inventory Fill rate increases if safety inventory is increased. This also increases the cycle service level. ► Lot size Fill rate increases with the increase of the lot size even though cycle service level does not change. 44 Factors Affecting Fill Rate -- Page 42 ► Safety inventory Fill rate increases if safety inventory is increased. This also increases the cycle service level. ss , ESC , fr ,CSL ESC = ss[1 Fs (ss / s L )] s Lfs (ss / s L ) fr = 1- ESC/Q ss = Fs1(CSL) s L ► Lot size Fill rate increases on increasing the lot size even though cycle service level does not change. fr = 1- ESC/Q CSL = F(ROP, DL, sL) is independent of Q 45 Evaluating Safety Inventory Given Desired Fill Rate If desired fill rate is fr = 0.975, how much safety inventory should be held? ESC = (1 - fr)Q = 250 Solve ss ss ESC = 250 = ss 1 Fs s L f s s L sL ss ss 250 = ss 1 NORMSDIST ,1,1,0 s NORMDIST L s s L L 46 Excel-Demo 47 臺灣大學 郭瑞祥老師 Evaluating Safety Inventory Given Desired Fill Rate If desired fill rate is fr = 0.975, how much safety inventory should be held? ESC = (1 - fr)Q = 250 Solve ss ss ESC = 250 = ss 1 Fs s L f s s L sL ss ss 250 = ss 1 NORMSDIST ,1,1,0 s NORMDIST L s s L L 48 Evaluating Safety Inventory Given Fill Rate The required safety inventory grows rapidly with an increase in the desired product availability (fill rate). 49 Fill Rate Safety Inventory 97.5% 67 98.0% 183 98.5% 321 99.0% 499 99.5% 767 Two Managerial Levers to Reduce Safety Inventory Safety inventory increases with an increase in the lead time and the standard deviation of periodic demand. ► Reduce the supplier lead time (L) – If lead time decreases by a factor of k, safety inventory in the retailer decreases by a factor of k . – It is important for the retailer to share some of the resulting benefits to the supplier. ► Reduce the underlying uncertainty of demand ( sD ) – If sD is reduced by a factor of k, safety inventory decreases by a factor of k. – The reduction in sD can be achieved by reducing forecast uncertainty, such as by sharing demand information through the supply chain. 50 Impact of Supply (Lead time) Uncertainty on Safety Inventory ► Assume demand per period and replenishment lead time are normally distributed D:Average demand per period sD:Standard deviation of demand per period (demand uncertainty) L: Average lead time for replenishment SL:Standard deviation of lead time (supply uncertainty) ► Consider continuous review policy, we have: Demand during the lead time is N(DL,sL2) DL = DL s 51 L s = L 2 D D 2SL2 Example ► Suppose we have D = 2,500 s D = 500 L = 7(days ) SL = 7(days ) CSL = 0.9 DL = DL = 2,500 7 = 17,500 s L s = L 2 D D 2SL2 = 7 5002 25002 72 = 17,550 Required safety inventory, ss = Fs1CSL s L = 22,491 ► A reduction in lead time uncertainty can help reduce safety inventory 52 SL σL 6 ss(days) 15,058 ss(units) 19,298 5 12,570 16,109 6.44 4 10,087 12,927 5.17 3 7,616 9,760 3.90 2 5,172 6,628 2.65 1 2,828 3,625 1.45 0 1,323 1,695 0.68 7.72 版權聲明 頁碼 53 作品 授權條件 作者/來源 5 本作品轉載自CoolCLIPS網站 (http://dir.coolclips.com/Popular/World_of_Industry/Food/Shopping_cart_full_of_gr oceries_vc012266.html ),瀏覽日期2011/12/28。依據著作權法第46、52、65條 合理使用。 6 本作品轉載自 WIKIPEDIA(http://en.wikipedia.org/wiki/File:Palm-m505.jpg),瀏 覽日期2012/2/21。 6 本作品轉載自Microsoft Office 2007多媒體藝廊,依據Microsoft 服務合約及著 作權法第46、52、65條合理使用。 19 臺灣大學 郭瑞祥老師 19 臺灣大學 郭瑞祥老師 22 臺灣大學 郭瑞祥老師 22 臺灣大學 郭瑞祥老師 版權聲明 頁碼 54 作品 授權條件 作者/來源 33 本作品轉載自 WIKIPEDIA(http://en.wikipedia.org/wiki/File:10_DM_Serie4_Vorderseite.jpg), 瀏覽日期2012/2/21。 43 臺灣大學 郭瑞祥老師 47 臺灣大學 郭瑞祥老師 47 臺灣大學 郭瑞祥老師