Inventory Management

profileHanlin
1617_ContinuousReview.pdf

1

A Continuous Review Inventory Model

2

Outline

 Types of Inventories  Pipeline stock, cycle stock, and safety stock

 A Continuous Review Inventory Model  System description

 Inventory concepts/equations

 Replenishment policy

 Performance measures

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

2

3

Related Readings

 Operations Management (10th Edition, Prentice Hall): part of Chapter 9, pages 319-325

 Operations Management (13th Edition, Pearson): part of Chapter 12, pages 508-512

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

4

Types of Inventories

 Pipeline stock (due to time and distance)  Things in transit from point A to point B

 Cycle stock (due to economies of scale)  Order in batches ==> “waves” of inventories

 Safety stock (due to uncertainty)  Inventories above the mean

A B

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

3

5

Managing Campbell’s chicken soup  How does Wal-Mart manage inventory for

Campbell's chicken soup?

 Consider the following inventory problem for Wal-

Mart:

 Demand rate is random over time.

 There is an inventory holding cost for each unit on hand

 There is an ordering/setup cost for each ordering/batch production

 Constant and deterministic supply lead-time

 Full backlogging

 There is a service level constraint (e.g., the probability of stock-out should be less than 5%)

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

6

System Description

customer demand

supplier

inventory system

 Assumptions  Supplier is reliable

 Supply lead time is constant, L periods

 Customer demand in a period is normally distributed with mean and variance

 Demands in different periods are independent

 When demand exceeds on-hand inventory, backlogging

 2

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

4

7

Important Inventory Concepts  Net inventory = on-hand - backorders

 Inventory position = net inventory + pipeline inventory

net inventory

inventory position

pipeline inventory

orders demand lead time = L

Observations:

(1) Net inventory is increased when a shipment is received and decreased when demand arrives.

(2) Inventory position is increased when an order is placed and decreased when demand arrives.

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

8

Important Inventory Concepts (Cont’d)  On-order inventory / pipeline inventory / scheduled receipts

= the number of units that have been ordered but have not been received.

 On-hand inventory = the number of units physically in inventory ready to serve demand.

 Backorder = the total amount of demand that has not been satisfied:

 All backordered demand is eventually filled, i.e., there are no lost sales.

 Inventory level / Net inventory = On-hand inventory - Backorder.

 Inventory position = On-order inventory + Inventory level.

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

5

9

An Example

In transit to store

Past orders of 4 units still need to be delivered

no backorders

In this example…

On hand = 6

Backlogs = 0 Inv. level = 6 – 0 = 6

Scheduled receipts = 3 + 4 = 7

X = Inventory Position = 6 + 7 = 13

The inventory position of the product (from the store’s perspective) is 13 units.

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

10

A Continuous Review Inventory Model

 A continuous review inventory system (also called Q- system) tracks the remaining inventory continuously to determine whether it is time to reorder.

 Whenever the inventory position drops to (or below) the level R, then place an order of Q units.

 The control policy involves two parameters (R, Q)

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

6

11

(R,Q) Policy Calculations

 Two parameters specify the policy: R and Q.

 The value of Q can be determined using the EOQ model

 If demand is certain, then

 If demand is uncertain, then

R = Mean Demand During Lead-time + SAFETY STOCK

R = Demand During Lead-time

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

12

Continuous Review with Certain Demand

Time

O n

-h a n

d i n

v e n

to ry

R

TBO

L

TBO

L

TBO

L

Order received

Order received

Q

OH

Order placed

IP

Order received

Q

OH

Order placed

IP

Order received

Order placed

IP

Q

OH

Note: This is essentially an EOQ model.

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

7

13

Continuous Review with Uncertain Demand

Time

O n

-h a n

d i n

v e n

to ry

TBO1 TBO2 TBO3

L L L

R

Order received

Q

Order placed

Order placed

Order received

IP IP

Q

Order placed

Q

Order received

Order received

OH

Note: R is the amount of inventory we will use to satisfy the lead time demand

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

14

Use Service Level to Determine R  Cycle service level (CSL, also called In-stock Probability)

measures the likelihood of not running into a stock-out by the end

of the lead time period.

CSL = Prob {demand during lead time ≤ R}

 Using statistics (for the case of normal distributions): z = NORMSINV(CSL),

NORMSINV is the inverse function of the standard normal distribution function

 Safety stock = zL  z = The number of standard deviations needed for a given

cycle-service level.  L = The standard deviation of demand during the lead time

period.

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

8

15

Scaling the Normal Distribution

N(=3192,=1181) N(=0,=1)

0 1250 2500 3750 5000

R -3 -1.75 -0.5 0.75 2

z

Scale down: z=(R-)/

Scale up: R = +z

Distribution function F(R) of any Normal distribution with mean  and st. dev.  can be found using the table of the Standard Normal distribution (z) with =0 and =1. That is, F(R) = (z) = ((R-)/). Thus we can look up z first and then convert z to R. In particular, we have

R = mean of lead time demand + z*(st. dev. of lead time demand)

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

16

A Normal Probability Distribution for an 85% In-stock Probability

Average demand

during lead time

Mean demand

during lead time

Cycle-service level = 85%

Probability of stock-out (1.0 – 0.85 = 0.15)

zL

R

Probability distribution of the lead-time demand

Let z satisfy NORMSDIST(z) = 0.85 and L be the standard deviation of lead-time demand, then safety stock = zL

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

9

17

Standard Normal Distribution Function Table

Examples:

NORMSDIST(2.33) = ?

NORMSINV(0.92) = ?

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

18

How to Obtain the Lead Time Demand Distribution

 Suppose the demand distribution in each period is normal with a

mean  and a standard deviation .

 The lead time is L periods and the demand distributions are

independent and identical across period.

 Then the distribution for the demand during the lead time has a

mean L = L and a standard deviation LL  

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

10

19

An Example

 = 15

+  = 75

Demand for week 1

L = 26

L = 225 Demand for 3-week lead time

+ 75

Demand for week 2

 = 15

= 75

Demand for week 3

 = 15

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

20

Calculating Total Costs in a Continuous Review Model

 Total policy-related costs for the continuous review system is the

sum of three cost components:

= Annual cycle inventory holding cost + annual ordering cost

+ annual safety stock holding cost

= LHzK

Q

D H

Q  )()(

2

where D: Mean annual demand Q: Economic order quantity K: Fixed ordering cost H: Annual inventory carrying cost

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

11

21

Example 12.4: Finding R and Total Costs Suppose that the average demand for bird feeders is 18 units per week with a standard deviation of 5 units. The lead time is constant at 2 weeks. Determine the safety stock and reorder point for a 90 percent cycle- service level. (Recall EOQ = 75, K = $45 per order, and H = $15 per unit per year from Example 12.2) What is the optimal R (rounded to the nearest integer)?

Answer:

L =  L = 5 2 = 7.1

Safety stock = zL = 1.29(7.1) = 9.16 or 9 units Reorder point = L + safety stock = 2(18) + 9 = 45 units

Demand distribution for lead time must be developed:

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

22

Example 12.4: Finding R and Total Costs What is the total policy-related cost of the Q system (keep two digits after the decimal point)?

Answer:

Safety stock = zL = 1.29(7.1) = 9.16 or 9 units Reorder point = L + safety stock = 2(18) + 9 = 45 units

C = ($15) + ($45) + 9($15) 75

2

936

75

C = $562.50 + $561.60 + $135 = $1259.10

(Recall Q = EOQ = 75)

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

12

23

Calculating Performance Measures for a (R,Q) Policy

 Suppose we use a (R,Q) inventory policy, with given parameters R and Q.

 What is the CSL this policy achieves?

 

  

  

L

LRNORMSDISTCSL  

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

24

Example: Finding (Q, R) and the CSL A regional warehouse purchases hand tools from various suppliers and then distributes to retailers in the region. The warehouse operates 5 days a week, 52 weeks per year. The following data are estimated for 3/8-inch hand drills with double insulation and variable speeds:

Average daily demand  = 100 drills Standard deviation of daily demand  = 30 drills Lead time L = 3 days

Holding cost H = $9.40 per unit per year

Ordering cost K = $35 per order

Cycle-service level = 92%

(a) What order quantity Q (rounded to the nearest integer) should be used?

(b) What reorder point R (rounded to the nearest integer) should be used?

(c) If on-hand inventory is 40 units, one order for 440 drills is pending, and no backorder exists, should a new order be placed?

(d) Suppose the manager has chosen R = 400. What will CSL be? Keep three digits after the decimal point.

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

13

25

Example: Solution a) Annual demand is D = 5*52 *100 = 26,000 drills a year

The order quantity is

b) The standard deviation for lead-time demand is

At 92% CSL, we have z = 1.41 (from Normal distribution Table). Therefore, safety stock = zL = 1.41*52  73 drills and reorder point R = 100(3) + 73 = 373 drills.

c) IP = OH +SR - BO = 40 + 440 -0 = 480 drills > R, so do not place a new order

d)

drills H

DK EOQ 440167,193

40.9$

)35)($000,26(22 

drillsLL 52330 

 

  

  

L

LRNORMSDISTCSL  

.973.0)923.1( 52

300400 

 

   

 NORMSDISTNORMSDIST

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang

26

Summary

 Inventory Concepts and Equations

 Continuous Review Inventory Models  Policy calculation for a given service level requirement

 Service level evaluation for a given control policy

Classes 16&17-Continuous Review Inventory Model MGT 303 Prof. Yang