Chapter 4: Individual Items with

Probabilistic Demand

 Characteristicsof stochastic demand

 The base-stock policy

 The (Q,r) or (Q,s) model
 The (R,r) or (s,S) model
 Periodic Review System – (S,T) model
Verify planned shortage case in Chapter 2
 1. Characteristics

 Demand per unit time is a random variable X with

mean E(X) and standard deviation σ don't know future demand
 Possibility of overstocking (excess inventory) or
understocking (shortages)
most of the time is
holding cost +
expiration value of
items

 There are overage costs for overstocking and

shortage costs for understocking
apply periodic
checking
routine: stationaries, clean service
Leverage: back-up generator for building, elevator, tv
bottleneck: raw items, package, bottle in coca cola
critical: the most important: machines, building, vehicle

apply regular and classify items into 4 groups: items are very important
continous which are without it the systems will stop
checking system
 Types of Stochastic Models
plan to meet a single period only
 Single period models
have to answer how much to order
apply when a company introduce new product to market/ when seasonal products
like moon cakes
 Fashion goods, perishable goods, goods with short lifecycles, seasonal goods
 One time decision (how much to order).
newsboy problems
 Multiple period models
Goods with recurring demand but whose demand varies from period to period
Inventory systems with periodic review
Periodic decisions (how much to order in each period)
 Continuous time models
Goods with recurring demand but with variable inter-arrival times between customer
orders

Inventory system with continuous review

Continuous decisions (continuously deciding on how much to order)

 Example
 If L is the order replenishment lead time, D is demand per unit
time, and r is the reorder point (in a continuous review system),
then:
Probability of stockout = 1- P(demand during lead time ≤ r)
If demand during lead time is normally distributed with mean
E(D)L, then choosing r = E(D)L leads to

Probability of stockout = 0.5

demand x leadtime
 29

2. The Base Stock Model

5
 Example
Consider the situation facing an appliance store sells a
particular model of TV. Because space is limited and
because the manufacturer makes frequent deliveries of
other appliances, the store finds it practical to order
replacement TVs each time one is sold. In fact, they
have a system that places purchase orders (POs)
automatically whenever a sale is made. But, because
the manufacturer is slow to fill replenishment orders,
the store must carry some stock in order to meet
customer demands promptly. Under these conditions,
the key question is how much stock to carry?
 The Base-Stock Policy
each time we order Q=1 , assume the lead time L is constant
we need to determine R, base stock level

In the base stock model, inventory is refilled one unit at a time and demand
is random. base stock level

Start with an initial amount of inventory R. Each time a new demand arrives, place
a replenishment order with the supplier.
An order placed with the supplier is delivered L units of time after it is placed.
Because demand is stochastic, we can have multiple orders (inventory on-order)
that have been placed but not delivered yet.
The amount of demand that arrives during the replenishment leadtime L is called
the leadtime demand (inventory on order).

7
 Assumptions
- possion distribution
indicate # customers
arrive in a period of time
- exponential distribution
indicate the time b/t
consecutive customers

1. Demands occur one at a time.

2. Any demand not filled from stock is backordered.
3. Replenishment lead times are fixed and known;
4. Times between consecutive orders are stochastic but
independent and identically distributed (i.i.d.)
5. Inventory is reviewed continuously
6. There is no fixed cost associated with placing an order; and
7. There is no constraint on the number of orders that can be
placed per year.
--> ordering cost is 0

 Last two assumptions imply that there is no incentive to

replenish stock in anything other than one-at-a-time fashion.,can
but it

lead
to
stock
out
 Notation
l = replenishment lead time , constant lead time
cumulative distribution
X = random demand during leadtime (pcs). function

 G(x) = Prob{X ≤ x}, c.d.f of demand during leadtime probability mass

 p(x) = Prob{X = x}, p.m.f of demand during leadtime function
  = E(X) mean demand during leadtime.
 = std dev of demand during leaditme.
h = unit holding cost
b = unit backorder cost
r = reorder point (pcs)
R = r +1, base stock level (pcs)
s = r - , safety stock level (pcs).r> mean during lead time
Q = 1, order quantity (fixed at one)
 S(R) = service level (average fill rate), fraction of orders filled from stock.
 B(R) = average backorder level orders met by inventory

 I(R) = average on-hand inventory level

 Base Stock Equations
 Under Base Stock Policy: inventory position = R (available stock)
Inventory position = on-hand - backorders + orders
 At any point in time, number of orders equals number demands during
leadtime(X) → on-hand inventory - backorders = R - X
 Expected Service Level: normal
distrbution
poison
distribu

 On-hand and backorders are never positive at the same time, so if X=x,
tion
discrete use sum
then: continuous use integral

probability that

 Expected backorder level:

demand>ROP

0, 𝑖𝑓 𝑥 < 𝑅
B R =𝜃− σ𝑅𝑥=0 1 − 𝐺(𝑥) = ቊ ; where:  = E(D)l
𝑥 − 𝑅, 𝑖𝑓 𝑥 ≥ 𝑅
 Expected inventory level: I(R) = R -  + B(R)
on-hand inventory
 Base Stock Example discrete case

l = one month

 = 10 units (per month)

Assume Poisson demand, so

Note: Poisson
demand is a good
choice when no
variability data
is available.
Base Stock Example Calculations
 Base Stock Example Results

 Service Level: For fill rate of 90%, we must set R-1= r =

14, so R=15 and safety stock s = r -  = 4. Resulting
service is 91.7%.

 Backorder Level:
B(r) = 0.187
 Inventory Level:

I(R) = R -  + B(R) = 15 - 10 + 0.187 = 5.187

 “Optimal” Base Stock Levels continuous case

don't know period in advance, but know the cost

 Objective Function:
TC(R) = hI(R) + bB(R) = h(R -  +B(R)) + bB(R)
→ TC(R) = h(R - ) + (h+b)B(R)
 Solution: if we assume G is continuous, we can use calculus to get
Implication: set base stock
level so fill rate is b/(h+b).
Note: R* increases in b and
decreases in h.
 If G is normal(, ), then
𝑏 𝑅∗ − 
𝐺 𝑅∗ = ↔Φ =Φ 𝑧
ℎ+𝑏 𝜎
where Φ(z)=b/(h+b) → z = NORMSINV(b/(h+b)}) POISSON.DIST

 So: R* =  + z 
 “Optimal” Base Stock Example
 Data: Approximate Poisson by normal with mean 10
units/month and standard deviation  = 3.16 units/month.
Set h=$15, b=$25.

 Calculations: normalinvr(G(R))
since Φ(0.32) = 0.625, z =0.32 and hence
R* =  + z = 10 + 0.32(3.16) = 11.01 ~ 11 r =10
 Observation: from previous table fill rate is G(10) = 0.583, so
maybe backorder cost is too low.
 Base Stock Insights
1. Reorder points control the probability of stockouts by
establishing safety stock.
2. The “optimal” fill rate is an increasing function of the backorder
cost and a decreasing function of the holding cost. We can
use either a service constraint or a backorder cost to
determine the appropriate base stock level.
3. Base stock levels in multi-stage production systems are very
similar to kanban systems and therefore the above insights
apply to those systems as well.
4. Base stock model allows us to quantify benefits of inventory
pooling.
3. The (Q,r) or (Q,s) Model
lead time is constant
 Assumptions
Demand occurs continuously over time
Times between consecutive orders are stochastic but independent and identically
distributed (i.i.d.)
Inventory is reviewed continuously
Supply leadtime is a fixed constant L
 Orders that cannot be fulfilled immediately from on-hand inventory are
backordered /shortage (lostsale)
Fixed cost associated with replenishment orders and cost per backorder.
Constraint on number of replenishment orders per year and service constraint

18
 The (Q, r) Policy
demand is Poisson or approximate normal distribution

Start with an initial amount of inventory R. When inventory level

reaches level r, place an order in the amount Q = R-r to bring
inventory position back up to level R. Thereafter whenever inventory
position drops to r, place an order of size Q.
The base-stock policy is the special case of the (Q, r) policy where Q
= 1.
+lost sale cost
Objective:

As in EOQ, this makes

batch production attractive.
19
(Q,r) Notation

(lost sale cost)

 (Q,r) Notation (cont.)
Decision Variables:

Performance Measures:
Inventory versus Time

2
 Demand during Leadtime

2
 Demand during Leadtime

r r)

2
 Costs in (Q,r) Model

 Fixed Setup Cost: AF(Q) expected lost sale

cost

 Stockout Cost: kD(1-S(Q,r)), where k is cost per stockout

 Backorder Cost: bB(Q,r)

 Inventory Carrying Costs: cI(Q,r)

 Fixed Setup Cost in (Q,r) Model

 Observation: since the number of orders per year

is D/Q,
 Stockout Cost in (Q,r) Model

 Key Observation: inventory position is uniformly distributed

between r+1 and r+Q. So, service in (Q,r) model is
weighted sum of service in base stock model.

 Result:

Note: this form is easier to use

in spreadsheets because it does
not involve a sum.
 Service Level Approximations

 Type I (base stock):

Note: computes number
of stockouts per cycle,
underestimates S(Q,r)

 Type II:
Note: neglects B(r,Q)
term, underestimates S(Q,r)
 Backorder Costs in (Q,r) Model

 Key Observation: B(Q,r) can also be computed by

averaging base stock backorder level function over the
range [r+1,r+Q].

 Result:

Notes:
1. B(Q,r)» B(r) is a base stock approximation for backorder level.

2. If we can compute B(x) (base stock backorder level function),

then we can compute stockout and backorder costs in (Q,r) model.
 Inventory Costs in (Q,r) Model

 Approximate Analysis: on average inventory declines from

Q+s to s+1 so

 Exact Analysis: this neglects backorders, which add to

average inventory since on-hand inventory can never go
below zero. The corrected version turns out to be
3.1. (Q,r) Model with Backorder Cost

 Objective Function:

 Approximation: B(Q,r) makes optimization complicated

because it depends on both Q and r. To simplify,
approximate with base stock backorder formula, B(r):
Results of Approximate Optimization
 Assumptions:
 Q,r can be treated as continuous variables
 G(x) is a continuous cdf

 Results:

Note: this is just the EOQ formula

Note: this is just the
base stock formula
if G is normal(q,s),
where F(z)=b/(h+b)

>> r = mean + s > s = z x sigma

 (Q,r) Example
D = 14 units per year
c = $150 per unit
h = 0.1 × 150 = $15 per unit
l = 45 days
 = (14 × 45)/365 = 1.726 units during replenishment lead time
A = $10
b = $40
Demand during lead time is Poisson
Values for Poisson(θ) Distribution
Calculations for Example
Performance Measures for Example
 Observations on Example

Orders placed at rate of 3.5 per year

Fill rate fairly high (90.4%)
Very few outstanding backorders (0.049 on
average)
Average on-hand inventory just below 3 (2.823)
 Varying the Example
 Change: suppose we order twice as often so F=7 per year, then
Q=2 and:

which may be too low, so increase r from 2 to 3:

3
 This is better. For this policy (Q=2, r=4) we can compute
B(2,3)=0.026, I(Q,r)=2.80.

 Conclusion: this has higher service and lower inventory than the
original policy (Q=4, r=2). But the cost of achieving this is an
extra 3.5 replenishment orders per year.
3.2. (Q,r) Model with Stockout Cost
= Type II/I

 Objective Function:

 Approximation: Assume we can still use EOQ to compute

Q* but replace S(Q,r) by Type II approximation and B(Q,r)
by base stock approximation:
Results of Approximate Optimization

 Assumptions:
Q,r can be treated as continuous variables
G(x) is a continuous cdf

 Results:

Note: this is just the EOQ formula

Note: another version

of base stock formula
if G is normal(q,s), (only z is different)
where F(z)=kD/(kD+hQ)
 Backorder vs. Stockout Model
 Backorder ModelDEPEND ON STOCK OUT TIME
 when real concern is about stockout time
 because B(Q,r) is proportional to time orders wait for backorders
 useful in multi-level systems

 Stockout Model Depend ON FILL RATE

 when concern is about fill rate

better approximation of lost sales situations (e.g., retail)

 Note:
 We can use either model to generate frontier of solutions
 Keep track of all performance measures regardless of model
 B-model will work best for backorders, S-model for stockouts
 3.3. Lead Time Variability
 Problem: replenishment lead times may be variable, which increases
variability of lead time demand.
 Notation:
L = replenishment lead time (days), a random variable
l = E[L] = expected replenishment lead time (days)
L = Var(L) = std dev of replenishment lead time (days)
D = demand per period t, a random variable, assumed i.i.d. demand rate
d = E(D) = expected demand per period
D = Var(D) = variance demand
 We have:

E(X)=E(L)E(D) demand during lead time

Var(X)=E(L)Var(D) + E(D)2Var(L)
Including Lead Time Variability in Formulas

 Standard Deviation of Lead Time Demand:

if demand is Poisson

Inflation term due to

lead time variability

 Modified Base Stock Formula (Poisson demand case):

Note: s can be used in any
base stock or (Q,r) formula
as before. In general, it will
inflate safety stock.
 Reorder Point with Lead Time Variability

Under the Normal approximation, the optimal reorder can be

obtained as

44
Expected Leadtime Demand

45
Variance of Leadtime Demand

46
Variance of Leadtime Demand

47
 Single Product (Q,r) Insights
 Basic Insights: invt position vs invt level:
 Safety stock provides a buffer against stockouts. - normally invt position > invt level
- invt level is on hand invt
 Cycle stock is an alternative to setups/orders.depend on Q - invt position = on hand + orders

 Other Insights:
1. Increasing D tends to increase optimal order quantity Q.
2. Increasing  tends to increase the optimal reorder point. (Note: either
increasing D or l increases .)
3. Increasing the variability of the demand process tends to increase the optimal
reorder point (provided z > 0).
4. Increasing the holding cost tends to decrease the optimal order quantity and
reorder point.
3.4. Uncertain Demand with Safety Stock
demand follow any distribution

IP
IP
Order
Order
Order received
received
Order
On-hand inventory

received
received
Q
Q Q
On
Hand
r
Order Order Order
placed placed placed

L1 L2 L3 Time
TBO1 TBO2 TBO3
49
 Expected Inventory (Assumptions)

I(t)

Q Slope
-D

SS Q
T=
D

Time
50
 Expected cost function
 Include expected: holding (h), setup (A), penalty (p – e.g.
backorder) and ordering (per unit) costs (c)
 Average Holding Cost:

 Q
h  SS + 
 Average Set-up Cost:
 2

A AD
=
T Q

51
 Expected cost function
⚫ Expected Shortage per Cycle:
¥
E(max(X - r, 0)) = ò (x - r) g(x) dx = n(r)
r

⚫ Interpret n(r) as the expected number of stockouts per cycle given

by the loss integral formula.
⚫ The unit normal loss integral values appear in Table B.1 (Gu(k)).

⚫ Expected Penalty Cost :

n (r ) D * n (r )
p =( )p
T Q
 Cost Minimization

⚫ Expected Cost Function:

Q D pDn(r)
Y (Q, r) = h( + r - Dℓ ) + A + + Dc
2 Q Q
⚫ Partial Derivatives:
(1)

¶Y h D pDn(r) 2D [ A + p n(r)]
= -A 2 - =0 => Q =
¶Q 2 Q Q 2
h

(2) This is the first equation

¶Y pD we will use to determine
= h+ n¢(r) optimal values Q and r
¶r Q
 Cost Minimization

⚫ Partial Derivatives:
(2)
¶Y pD
= h+ n¢(r)
¶r Q
¥
Note: n(r) = ò (x - r)g(x)dx
r

n¢(r) = -(1- G(r))

Qh
1- G(r) = This is the second equation
pD we will use to determine
optimal values Q and r
 Solution Procedure
when Q and r do
not change any
more

⚫ The optimal solution procedure requires iterating between

the two equations for Q and r until convergence occurs

⚫ A cost effective approximation is to set Q=EOQ and find

R from the second equation.
 Finding Q and r, iteratively
1. Compute Q = EOQ.
2. Substitute Q in to Equation (2) and compute r.
Qh
1- G(r) =
pD
3. Use r to compute n(r) in Equation (1).
¥
n(r) = ò (x - r)g(x)dx
r
4. Solve for Q in Equation (1).
2D [ A + p n(r)]
Q=
h
5. Go back to Step 2, continue until convergence.
 Example: Uniform Distribution
⚫ A company purchases air filters at a rate of 800 per year

⚫ $10 to place an order

⚫ Unit cost is $25 per filter

⚫ Inventory carry cost is $2/unit per year

⚫ Shortage cost is $5

⚫ Lead time is 2 weeks

⚫ Assume demand during lead time follows a uniform distribution from 0

to 200
⚫ Find (Q,r)
 Solution

⚫ Partial derivative outcomes:

2 D éë A + p n ( r )ùû 2(800)(10 + 5 n ( r ))
Q= =
h 2
= 8000 + 4000 n (r )
Qh 2Q Q
1- G ( r ) = = =
pD 5(800) 2000
58
 Solution
⚫ From Uniform U(0,200) distribution:

1 1
U(0,200): g(x) = =
b-a 200
¥ 200
n(r) = ò (x - r)g(x)dx = ò
1
(x - r) dx
r r 200
1 æç x 2 ö 1 æ 200 2
x=200
r 2 ö
= - rx ÷= ç - 200r - + r ÷
2

200 çè 2 x=r
÷ 200 è 2
ø 2 ø
r2
= 100 + -r
400
59
 r2
Solution n(r) =
400
- r +100

Qi = 8000 + 4000n(r)
Q
⚫ Iteration 1: 1- G(r) =
2000

2AD 2(10)(800)
EOQ = = = 8000 = 89.44 = Qo
h 2
G(r)
Qo h 89
1- G(ro ) = = = .04
pD 2000
G(ro ) = .96
0 200
ro = (.96)(200 - 0) = 192
R
60
 r2
Solution n(r) =
400
- r +100

Qi = 8000 + 4000n(r)
Q
⚫ Iteration 2: 1- G(r) =
2000

(192)2
n(r0 ) = -192 +100 = .198
400
Q1 = 8000 + 4000(.198) = 93.76
94
1- G(r1 ) = = .05
2000
Þ r1 = (.95)(200) = 190

61
 r2
Solution n(r) =
400
- r +100

Qi = 8000 + 4000n(r)
Q
⚫ Iteration 3: 1- G(r) =
2000

190 2
n(r1 ) = -190 +100 = .2197
400
Q2 = 8000 + 4000(.2197) = 94.228
94
(1- G(r2 )) = = .05
2000
r2 = 190

62
 Solution
⚫ r didn’t change => CONVERGENCE
⚫ (Q*,r*) = (94,190)

I(t)
253
Slope
190 -800

With lead time equal to 2 weeks:

159
SS = r – Dl =190-800(2/52)=159

63
 Example: Normal Distribution
⚫ Demand is Normally distributed with mean of 40 per week and
a weekly variance of 8

⚫ The ordering cost is $50

⚫ Lead time is two weeks

⚫ Shortages cost an estimated $5 per unit short to expedite

orders to appease customers

⚫ The holding cost is $0.0225 per week

⚫ Find (Q,r)

64
 Solution
⚫ Demand is N (40,2 2 ) per week.
⚫ Lead time is two weeks long. Thus, during the lead
time:
⚫ Mean demand is 2(40) = 80
⚫ Variance is (2*8) = 16
⚫ Demand observed in one week is independent from
demand observed in any other week:
⚫ E(demand over 2 weeks) = E (2*demand over week 1)
= 2 E(demand in a single week) = 2 μ = 80

Standard deviation over 2 weeks is σ = (2*8)0.5 = 4

65
 Solution
⚫ Iteration 1:
2AD 2(50)(40)
EOQ = = = 421.6 = Qo
h .0225
Qo h 421.6(.0225)
1- G(ro ) = = = .0473
pD 5(40)
G(ro ) = .9527
⚫ From the standard normal table:
G (ro ) = .9527
P ( z £ 1.67) = 0.9527
ro = m + z s = 80 +1.67(4) = 86.68
66
 Solution
This is the unit normal loss function L.
⚫ Iteration 2:
¥
n ( r0 ) = ò ( x - r ) g ( x ) dx
r
1 æ x -m ö
2
¥ - ç ÷
n ( r0 ) = ò ( x - r )
1 2è s ø
e dx
r 2ps
ær -mö æ 86.68 - 80 ö
=s Lç ÷ = 4L ç ÷ = 4 L (1.67)
è s ø è 4 ø
= 4(.0197) = .0788
67
 Solution

⚫Iteration 2:
n(r0 ) = .0788
2D ( A + pn(r)) 2(40) ( 50 + 5(.0788))
Q1 = = = 423.3
h .0225
Q1h 423.3(.0225)
1- G(r1 ) = = = .0476
pD 5(40)
G(r1 ) = .9523
r1 = 86.68
Convergence!
68
 3.5. Service Levels in (Q,r) Systems
cannot estimate penalty

 In many circumstances, the penalty cost, p, is difficult to

estimate.
 For this reason, it is common business practice to set inventory
levels to meet a specified service objective instead.
1) Type 1 service: Choose r so that the probability of not stocking
out in the lead time is equal to a specified value. Poisson
 Appropriate when a shortage occurrence has the same consequence
independent of its time and amount.
2) Type 2 service: Choose both Q and r so that the proportion of
demands satisfied from stock equals a specified value.
 In general,  is interpreted as the fill rate.
general case
know fill rate in advance
3.5.1. (Q,r) Systems with Type 1 Service Constraint

⚫For type 1 service, if the desired service level is

α then one finds r from G(r)= α and Q=EOQ
⚫Specify a, which is the proportion of cycles in
which no stockouts occur.
⚫ This is equal to the probability that demand is
satisfied.
G(r) = a ® probability demand is satisfied
2AD
Set Q = EOQ =
h

70
 3.5.2. (Q,r) Systems with Type 2 Service Constraint

 Type 2 service requires a complex iterative solution

procedure to find the best Q and r

 However, setting Q=EOQ and finding r to satisfy n(r) = (1-

)Q will generally give good results
expected #
shortage

Average Stockouts per Cycle n(r) n(r)

= =
Average Demand per Cycle DT Q

n( r )
= 1−  ,
Q
n(r ) = (1 −  )Q
71
 Type 2 Service Constraint
 May specify fill rate , and use EOQ for Q to compute r
 Or, solve for p (penalty caused by shortage) : 1- G ( r ) = Qh
pD
and substitute into the equation: æ Qhn(r) ö
2D ç A + ÷
2D ( A + pn(r)) è (1- G(r))D ø
Q= =
h h
hQ 2 hn(r)
= A+ Q
2D (1- G(r))D
hQ 2 hn(r)
- Q- A = 0
2D (1- G(r))D
 Result: n(r) 2AD æ n(r) ö
2

Q= + +ç ÷
1- G(r) h è 1- G(r) ø
n(r) = (1- b )Q
 4. The (R, r) Model
This is usually called the (s, S) model
Each demand order can be for multiple units
Demand orders are stochastic
A replenishment order is placed to bring inventory position back to
R
Decision variables are R (instead of Q) and r

73
 4.1. Order-Up-To-Level (s, S) vs (Q, r) System
(s,S) model have more invt than (Q,r)
(s,S) aplly min max rule

SS • If all demand
transactions are
unit-sized, two
systems are the
ss same S= r+Q
• (s,S)~ “min-max”
(s,S) (s,
system
S) system rR+Q
+Q

rR

(Q,,R)
(Q,r) system
system
74
 Alternative (s, S) Policy

 Define two levels, s < S, and let u be the starting inventory at

the beginning of a period.
 Then, if u is less or equal to s, order (S – u),
if u is more than s, do not place an order.
 In general, computing the optimal values of s and S is
extremely difficult, so few systems operate using optimal
(s,S) values.
 Approximation: Compute optimal values for (Q,r) model and
set
s = r and S = r + Q

75
4.2. Periodic Review (S, T) System
cycle time is fixed--> holding
the riskcost is higher
of stock out is higher because it depend on demand during leadtime and
cycle time

T T

76
 Periodic Review Systems
⚫ Continuous Review Systems
⚫ Always knew level of on-hand inventory
⚫ Could place an order at any time

⚫ Often, we are constrained by WHEN we can order -- and it

may be periodically
⚫ Train dispatched once a week
⚫ Delivery truck arrives each morning
can be applied for routine products

⚫ Thus, we do not need to continuously review inventory,

just check periodically

77
 Periodic Review Systems
⚫ If demand were known and constant, we would just resort
to our EOQ solution, possibly modifying it to meet
shipping date from EOQ can determine N=D/Q--> T=1/N--> adjust T to meet shipping
⚫ Now: demand is random variable
⚫ Setting:
⚫ Place an order every T periods
⚫ Policy: Order up to S
⚫ The value of Q (order quantity) will now change periodically
⚫ Previous concern: demand exceeding supply during the
lead time
⚫ Now: demand exceeding supply during the period and lead
time, or T + l

78
 Periodic Review Systems
Order up to S
every T periods of
time.
I(t) Order arrives.
S Cycle continues.
Demand
Q

Lead Time passes…

l l
Time
T
79
 Expected Cost Function
⚫ Include expected holding, setup, penalty
and ordering (per unit) costs
⚫ Average Inventory Level:
S
At level r*,on average,
order Q = S-r* units.
r* l periods later, units arrive.
Inventory level?
l

T
80
 Expected Cost Function
⚫ Include expected: holding, setup, penalty
and ordering (per unit) costs
⚫ Average Inventory Level:
S
S-Dl units present when
Q arrive (expected) as
Dl units consumed over
leading time.

81
 Expected Cost Function
⚫ Include expected: holding, setup, penalty
and ordering (per unit) costs
⚫ Average Inventory Level:
S S - Dl
D*T units
removed
(expected) from
inventory over
time T.
T

82
 Expected Cost Function
⚫ Include expected: holding, setup, penalty
and ordering (per unit) costs
⚫ Average Inventory Level:
S D*L S-Dl

D*T/2 D*T

S-Dl-D*T
T

DT DT
Average Inventory Level = ( S - Dℓ - DT ) + = S - Dℓ -
83
2 2
 Expected Cost Function
⚫ Include expected:
holding, setup, penalty and ordering (per unit) costs

⚫ Average Holding Cost:

æ DT ö
h ç S - Dℓ - ÷
è 2 ø

⚫ Average Set-up Cost: A

T
84
 Expected Cost Function
⚫ Expected Shortage per Cycle:
⚫ f(x)dx = P(demand in T + l is between x and x + dx)
if demand during cycle time and lead time > target level --> shortage

E (max( D (T + ℓ ) - S ,0))
¥
= ò ( x - S ) g ( x ) dx = n ( S )
S

⚫ Expected Penalty Cost :

n(S)
p
T
85
 Cost Minimization
⚫ Expected Cost Function:

DT A pn(S)
Y (S,T ) = h(S - Dℓ - )+ + + Dc
2 T T

⚫ Derivative:
⚫ Recall that T and l are given:

dY p
= h + n¢(S)
dS T
86
 Cost Minimization
⚫ Derivative:
dY p
= h + n¢(S) = 0
dS T
¥
Note: n(S) = ò (x - S)g(x)dx
S

n¢(S) = -(1- G(S))

hT
Þ 1- G(S) =
p
⚫ or:
p - hT
G(S) =
p
87
 Example
 A special control board is used in a version of a product on
the production line
 The board cost is $122.50
 The holding cost rate is 30% per year
 Reorders are placed at the start of each week, and the
supplier delivers these parts in one week
 The shortage cost is $100 per board due to worker
downtime
 Weekly demand is N (μ=125, σ2=104.17)
 Set up cost (A) is $120
 Find S

88
 Solution
⚫ Holding cost is:
h = Ic = .30 (122.50) = 36.75 / 52 = $.7067 per week
⚫ Compute:
p - hT 100 - (.7067)1
G(S) = = = 0.993
p 100
⚫ Demand Distribution is Normal
⚫ mean = 125 variance = 104.17
⚫ Z = 2.455 from Normal table
⚫ S =  + z → S = 125+(2.455)(104.17)1/2 = 150.06

89
 Solution

⚫ If penalty cost drops to $10 per unit:

⚫ Compute:
p - hT 10 - (.7067)1
G(S) = = = 0.929
p 10
⚫S = 125+(1.47)(104.17)1/2 = 140
⚫ p = 1? S = 125+(-.54)(104.17)1/2 = 119.49

90