Int. J.

Production Economics 155 (2014) 222228

Contents lists available at ScienceDirect

Int. J. Production Economics

journal homepage: www.elsevier.com/locate/ijpe

Optimal vendor investment for reducing defect rate in a vendorbuyer

integrated system with imperfect production process
O. Dey n, B.C. Giri
Department of Mathematics, Jadavpur University, Kolkata 700032, India

art ic l e i nf o

a b s t r a c t

Article history:
Received 29 May 2013
Accepted 4 February 2014
Available online 11 February 2014

This paper investigates a single-vendor single-buyer integrated production-inventory model with

stochastic demand and imperfect production process. It is assumed that there is an inspection activity
on the part of the buyer with a xed screening rate greater than the demand rate. The vendor invests
money in order to improve the production process quality and reduce the number of defectives. The
expected annual integrated total cost is derived under the n-shipment policy and an iterative procedure
is suggested to determine the optimal decisions. The benet of investment in reducing the defect rate is
illustrated by way of numerical examples. It is observed that, the higher the defect rate, the more
benecial the investment. Numerical studies further explore that an increased demand requires an
increased investment to optimize the total cost.
& 2014 Elsevier B.V. All rights reserved.

Keywords:
Inventory
Production
Economic order quantity
Vendorbuyer
Defective items
Inspection
Process quality

1. Introduction
Inventory management is one of the key components of any
business that can be controlled by a business manager to efciently and successfully operate in the ercely competitive modern
global market. Therefore, it is very important to build inventory
models that are sensitive and responsive to the dynamic real life
market situations. The economic order quantity (EOQ), rst
proposed by Ford Whitman Harris (1913), is the most fundamental
result which has generated whole new directions of research in
inventory management since its inception.
Integration of vendor's and buyer's individual problems in a
supply chain has been a point of interest of many supply chain
researchers during the past few decades. This is because integrated policy has the ability to offer customers shorter lead-time
and lower inventory cost. It also helps to determine problem areas
along the process enabling businesses to take decisive action and
further reduce cost to improve the nal price. Improved customer
satisfaction and loyalty is a byproduct of an integrated supply
chain because the end customers experience improved on-time
delivery. It also makes the system as a whole more robust enabling
both the vendor and the buyer to be more exible in dealing with
sudden disruptions.

Corresponding author.
E-mail addresses: oshmi_kgp@yahoo.co.uk (O. Dey),
bibhas_pnu@yahoo.com (B.C. Giri).
http://dx.doi.org/10.1016/j.ijpe.2014.02.004
0925-5273 & 2014 Elsevier B.V. All rights reserved.

In the literature of single-vendor single-buyer integrated

inventory model, it is often assumed that the demand is deterministic and shortages are not allowed. Ben-Daya and Hariga (2004)
extended this by taking the annual customer demand to be
stochastic and thereby allowing shortages. The production process
was however taken to be perfect. Even though production process
is often considered to be perfect, but in reality, it is extremely
unlikely that a production process is 100% defect-free. Even in
models that do consider imperfect production process (Huang,
2004), the process quality is not assumed to be a control
parameter. Porteus (1986) suggested that the process quality can
be improved by making an investment in the production process
in terms of buying new equipment, improving machine maintenance and repair, worker training, etc. Also, controlling process
quality is an important tool in the hands of the decision maker as
it is expected to lead to the production of smaller batch sizes of
better quality products. In an integrated model, since the production is controlled by the vendor who has to pay warranty cost for
defective items, it is benecial to him, in particular, and to the
supply chain as a whole, to invest in reducing the number of
defective items produced. This control of process quality by the
vendor affects production yield rate of non-defective items which,
in turn, inuences other important decisions such as the vendor's
production lot size and the number of shipments delivered from
the vendor to buyer. Evidently these decisions are directly related
to the total cost incurred by the supply chain as a whole.
Consequently, the process quality and the optimal policy need to
be determined jointly in order to minimize the total cost associated with the supply chain. This also provides added advantages

O. Dey, B.C. Giri / Int. J. Production Economics 155 (2014) 222228

of reduced warranty cost and perhaps most importantly provides

better quality products to the buyer. Also, a buyer is more likely to
place an order to a vendor who has a reputation for producing
better quality items. To incorporate the issue of quality improvement, it is assumed that the vendor makes an investment to
reduce the defect rate of the production process.
For integrated models with imperfect production process, it is
very likely that the buyer performs some sort of inspection activity
before selling the products to the customers. Ignoring this inspection/screening period or assuming it to be negligible is not very
practical. Therefore, we assume that the buyer performs an errorfree and non-destructive screening in a non-negligible nite period.
At the end of the screening period, all the defective items in each lot
are returned to the vendor at the time of the next delivery.
This paper, therefore, extends the existing literature by investigating an integrated single-vendor single-buyer productioninventory model with stochastic demand, an imperfect production
process and a non-negligible nite screening period. Since it is
assumed that the vendor makes an investment in improving the
process quality, therefore, the defect rate is assumed to be an
additional control parameter together with the number of shipments from the vendor to the buyer, the safety stock factor and the
buyer's order quantity.
The rest of the paper is organized as follows: Section 2 presents
a brief review of related literature. The proposed model is
formulated in Section 3 and the solution procedure is outlined in
Section 4. Section 5 illustrates the developed model with numerical examples. The paper is concluded in Section 6 with some
remarks and future research directions.

2. Literature review
In the recent past, several researchers have shown that both
the buyer and the vendor can better achieve their goals through
strategic cooperation with each other. Goyal (1976) was the rst to
develop an integrated inventory model for a single-supplier
single-buyer problem. Banerjee (1986) generalized Goyal's model
and developed a joint economic lot-size model where the vendor
produces on a lot-for-lot basis to meet the buyer's order under
deterministic conditions. Goyal (1988) extended Banerjee's model
suggesting that the vendor's economic production quantity per
cycle should be a positive integer multiple of the buyer's purchase
quantity. Since then a lot of research (Ha and Kim, 1997; Hill, 1997,
1999; Pan and Yang, 2002; Ouyang et al., 2006) has been devoted
to the study of integrated vendorbuyer model under various
assumptions. Recently researchers such as Ben-Daya and Hariga
(2004), Hsiao (2008), and Glock (2009, 2012) have also made some
advances in the integrated model under stochastic demand.
However, most of them did not consider the issue of process
quality and its relation to the production shipment schedule in
terms of the number and the size of batches transferred from the
vendor to the buyer.
Porteus (1986) was the rst to incorporate the concept of
imperfect production process into a basic EOQ model. He introduced
the concept of making an investment to improve the process quality
in terms of reducing the probability of the production process going
out-of-control. Lee and Rosenblatt (1987) considered process inspection during production for detecting and restoring out-of-control
systems. Researchers such as Rosenblatt and Lee (1986), Scwaller
(1988), Ben-Daya and Hariga (2000) and the references therein have
also addressed the issue of imperfect production in inventory
models. Goyal and Crdenas-Barrn (2002) proposed a practical
approach to the economic production quantity (EPQ) model with
imperfect items. A majority of these researches, however, focused on
determining the optimal policy solely from the buyer's or the

223

vendor's point of view. Zhang and Gerchak (1990) considered a

joint lot sizing policy and inspection policy in an EOQ model with
imperfect items. Salameh and Jaber (2000) investigated a joint lot
sizing and shipment policy assuming a random percentage of units
to be defective and the defective items be sold at the end of the
screening period as a single batch. Huang (2004) developed an
integrated vendorbuyer inventory model for items with imperfect
quality and equal shipment size in a deterministic framework,
assuming that the number of defective items followed a given
probability density function. He also assumed that the vendor
treated the defective items as a single batch at the end of the
buyer's 100% screening process. Shortages or any investment was
however not considered. Ouyang et al. (2006) investigated an
integrated model with imperfect production but did not consider
any investment, re-order point or shortages. Lin (2013) assumed
that, on the arrival of an order, the buyer performed a nondestructive and error-free screening process. The screening rate
was xed and the process was gradual so as to render the screening
time non-negligible. At the end of the screening period, all the
defective items in each lot were discovered and returned to the
vendor at the time of the next delivery. This led to the buyer having
two types of holding cost for defective and non-defective items.
However, in any of the above mentioned papers, the production
process quality was not treated as a control parameter and any
investment on part of the vendor to improve the production process
was not considered. Ouyang et al. (2007) developed a stochastic
integrated model with investment in process quality but neglected
the length of the inspection period. Shu and Zhou (2014) proposed
an integrated single-vendor single-buyer model in which the
products are sold with free minimal repair warranty. The integrated
total cost is minimized by optimizing the number of shipments, the
shipment quantity, the setup cost and the process quality. However,
in this paper the demand is assumed to be deterministic and
shortages are not allowed. Moreover, there is no consideration of
inspection/screening on part of the buyer.
Keeping these various issues in mind, the existing literature is
extended in this paper by considering an integrated single-vendor
single-buyer production-inventory model with stochastic demand,
a non-negligible nite screening period and investment to
improve the production process quality. For better understanding,
a comparison of the proposed model with some of the related
works in the literature is given in Table 1.
3. Model development
3.1. Notations and assumptions
To develop the proposed model, the following notations are
used:
D
P
A
B

expected demand rate (units/time) non-defective items

production rate (P=1/p)
buyer's ordering cost per order
vendors setup cost

Table 1
A comparison of the present model with some related works in the literature.
Paper

Demand

Production Screening Investment

Huang (2004)
Constant
Imperfect
Ben-Daya and Hariga (2004) Stochastic Perfect
Ouyang et al. (2007)
Stochastic Imperfect
Lin (2013)
Stochastic Imperfect
Shu and Zhou (2014)
Constant
Imperfect
Present model
Stochastic Imperfect

Yes
No
No
Yes
No
Yes

No
No
Yes
No
Yes
Yes

224

O. Dey, B.C. Giri / Int. J. Production Economics 155 (2014) 222228

F
K
L
hv
hb1
hb2
s
x
w
y

transportation cost per delivery

vendor's setup cost
lead-time
vendor's holding cost per item per year
buyer's holding cost for defective items per item per year
buyers holding cost for non-defective items per item per
year
buyer's unit screening cost
buyer's screening rate
vendor's unit warranty cost for defective items
percentage of defective items produced
buyer's shortage cost per item per year
fractional opportunity cost
percentage decrease in defective items per dollar increase
in investment

We develop the model with the following assumptions:

A single buyer orders items of a single product from a single

vendor.

Demand per unit time is normally distributed with mean D and

standard deviation s.

The buyer places an order of nQ (non-defective) items to the

vendor. The vendor produces these items and, on average,

transfers these items to the buyer in n equal sized shipments,
where n is a positive integer.
The buyer follows the classical Q ; r continuous review inventory policy with deterministic constant lead-time.
Lead-time L is a constant. The lead-time demand ispnormally

distributed with mean DL and standard deviation s L.

The re-order point r expectedpdemand
during lead-time saf
ety stock (SS) i.e., r DL ks L, where k is the safety stock
factor.
Shortages are allowed and completely backlogged.
y0 o y o1 is the percentage of defective items produced in
each batch of size Q.
The vendor's rate of production of non-defective items is
greater than the demand rate i.e., P1  y 4 D.
The screening rate x is xed and is greater than the demand
rate i.e., x 4 D.
The vendor incurs a warranty cost for each defective item
produced.
The vendor invests money to improve the production process
quality in terms of buying new equipment, improving machine
maintenance and repair, worker training, etc. We consider the
following logarithmic investment function Iy (Porteus, 1986):
 
1
y
Iy ln 0

y
where is the percentage decrease in y per dollar (or any other
suitable currency) increase in investment and y0 is the original
percentage of defective items produced prior to investment.

We suppose that the buyer places an order of size nQ for nondefective items to the vendor. In order to reduce the production
cost, the vendor produces these nQ items at one go and transfers n
batches of Q items each at regular intervals of Q 1  y=D units of
time on average. The length of each ordering cycle is therefore
Q 1  y=D and the length of the complete production cycle is
nQ 1  y=D.

order point r, the buyer places an order of size Q for non-defective

items to the vendor (Fig. 1). The vendor delivers these items after a
constant lead-time L. By assumption, since the safety factor k is
related to the re-order point, k is considered a decision variable
instead of r. On arrival of the order, the buyer inspects the items at
a xed screening rate x. It is assumed that the screening process is
non-destructive and error-free. The defective items in each lot are
discovered gradually, kept in hold separately and returned to the
vendor on the arrival of the next lot. The buyer, therefore, has two
types of holding cost for defective items and non-defective
items. The buyer's average inventory level for non-defective items
(including those defective items which have not yet been detected
before the end of the screening time Q =x) is given by
 p

nQ 1  y
Q 1  y
DQy
ks L

1
D
2
2x1  y
Similarly, the average inventory level for defective items is given
by


1y 1
nQ 2 y

2
D
2x
The annual expected total cost for the buyer including the
ordering cost, shipment cost, holding cost, shortage cost and
screening cost is, therefore, given by


DA nF
DQy
ETCBQ ; k; n
hb1 Qy 
nQ 1  y
2x1  y
 p

Q 1  y
DQy

hb2 ks L
2
2x1  y
p
Ds L k sD

3
Q 1  y
1y
R1
where k k z  kz dz, z being the standard normal
density function.
Now, if the buyer follows a deterministic demand, places his
order only when the inventory level falls to zero and receives the
order instantaneously then we have k 0; L 0; 0. Also, if the
buyer screens the items received as a single batch and rejects
those as a single batch at the end of the screening process then we
have hb1 0. Under these assumptions, the above expression for
the annual expected total cost for the buyer modies to


DA nF
Q 1  y
DQy
sD
hb2

4
ETCB
nQ 1  y
2
x1  y
1y
which is as given in Huang (2004).
If the production process is assumed to be perfect, i.e., y 0,
then the annual expected total cost for the buyer given in (3)
reduces to
p


p Q
DA nF
Ds L k
ETCB
hb2 ks L

5
nQ
2
Q
which is the same as in Ben-Daya and Hariga (2004), assuming the
lead-time L to be a known constant.

Q/ x
L

3.2. Buyer's perspective

From the buyer's point of view, it is assumed that as soon as the
inventory of non-defective items reaches the level called the re-

Q(1y) / D

Fig. 1. Inventory of the buyer.

O. Dey, B.C. Giri / Int. J. Production Economics 155 (2014) 222228

3.3. Vendor's perspective

During the production process, the vendor produces Q items in
the rst instance and delivers those to the buyer. After that, the
vendor delivers a quantity Q to the buyer every T units of time
where T Q 1  y=D. This continues till the vendor's production
run is completed (Fig. 2). By assumptions, the vendor's production
rate for non-defective items is greater than the demand rate.
Therefore, the vendor's inventory level gradually increases till the
production is over. Then the vendor's average inventory holding
cost (see Fig. 3 for holding area) can be calculated as follows
(Huang, 2004):
 


Q
Dp
2Dp
EHCV hv n 1 
1
6
2
1 y
1y
The annual expected total cost incurred by the vendor is, therefore, obtained as the sum of the setup cost, holding cost and
warranty cost for the defective items (Huang, 2004):
 


BD
Q
Dp
2Dp
wDy
ETCVQ ; n
hv n 1 
1

7
nQ 1  y
2
1y
1y
1y
Putting y 0 in (7), we get
BD
Q
hv n1  Dp 1 2Dp
ETCV
nQ
2

nQ / P

Fig. 2. Inventory of the vendor.

hands of the decision maker since it's control is needed to lower

associated costs incurred and the production of smaller batch sizes
of better quality products. Therefore, it is quite appropriate for the
vendor to make an investment to try and reduce the number of
defective items produced.
Assuming a logarithmic investment function of the form
Iy 1= lny0 =y, the expected annual total cost of the vendor
can be obtained as
 


BD
Q
Dp
2Dp
hv n 1 
1
ETCVQ ; y; n
nQ 1  y
2
1y
1y
 
wDy
y0

ln
9
1 y
y
where is the fractional opportunity cost.
It may be noted here that this logarithmic function Iy is a
convex function with respect to y for all y0 o y r y0 o 1. In the
later part of this paper, we will show that the characteristics of the
model obtained for the logarithmic investment function remain
essentially unaffected if another investment function viz., a sublinear power investment function is used instead.
3.4. Integrated approach

which is as given in Ben-Daya and Hariga (2004).

The total cost in (7) does not include any investment on the
part of the vendor to improve the process quality. However, as
mentioned earlier, process quality is an important tool in the

Q (1y) /D

225

The expected annual total cost of the integrated system is the

sum of the vendor's and the buyer's expected annual total costs
which is given by


DA B nF
DQy
ETCQ ; y; k; n
hb1 Qy
nQ 1  y
2x1  y
 


Q
Dp
2Dp
hv n 1 
1
2
1y
1y


p Q 1  y
DQy

hb2 ks L
2
2x1  y
p
 
Ds L k s wyD
y

ln 0
10

Q 1  y
1y
y

In the above objective function, the control parameters are Q ; y; k

and n. It is very difcult to show analytically that ETC is a convex
function in all the decision variables Q ; y; k; n. However, it can be
veried numerically that for given values of n (positive integer)
and y0 o yr y0 o 1, the total cost function ETC is convex in Q and
k. One instance of 3D-graph of ETC is shown in Fig. 5. In the
following section, we develop a solution procedure to derive the
optimal values of Q ; y; k and n such that the joint expected annual
total cost ETC is minimized.

4. Solution procedure
The total cost function ETC is convex in n, since it is easy to see
that
2 ETC
2dA K
40
3
n2
n Q 1  y

8 nZ1

11

For xed values of n and y0 o y r y0 o 1, ETC can also be shown

to be convex in k and Q, since
p
2 ETC d Lsk
40
12

2
Q 1  y
k
2 ETC
Q 2

Fig. 3. Vendor's inventory holding area.

2dGn
Q 3 1  y

p
2d Lsk
Q 3 1 y

40

13

where Gn A B nF=n.
It is, however, not possible to prove conclusively that ETC is
convex in y even though y is bounded. To derive an optimal
solution, we suggest the following procedure:

226

O. Dey, B.C. Giri / Int. J. Production Economics 155 (2014) 222228

For xed n, let us take the rst derivative of ETC with respect to
k and put it to zero. That is,
ETC
D
hb2
Fk 1 0
k
Q 1 y

14

where F is the cumulative distribution function.

This implies
F k

hb2 Q 1  y
D

15

where F  1  F.
Next, equating to zero the rst derivatives of ETC with respect
to Q and y, we have


ETC
DGn
D
 2
yhb1 1 
Q
2x1  y
Q 1  y


1y
Dy

hb2
2
2x1 y




hv
Dp
2Dp

1n 1
1y
1 y
2
p
Ds L k
0
16

Q 2 1  y


ETC
Dw Dt wy
DGn
D


Qhb1 1 
y
1  y 1  y2
2x1  y
y Q 1 y2
(
)
DQyhb1
Q
DQ
DQy

h


b2
2 2x1 y 1  y2
2x1  y2
(
)
p
Qh
2Dp
Dnp
Ds L k

0

v
2 1  y2 1  y2
Q 1  y2
On simplication, (16) yields the following:
s
p
DGn Ds L k
Q
Hn; y

ETC
dy

40
w
1 y

log

y0
y

20

o0

ETC

40
y0
y
17

18

(
)


Dy
1  y2 Dy
hb2

Hn; y hb1 y1  y 
2x
2x
2

hv
 n  2Dp n 11  y
2

It is to be mentioned here that the solution procedure adopted

above only provides a local optimum. As it is difcult to prove
analytically that the objective function in (10) is convex in all
control parameters, we cannot claim that the solution obtained
provides a global optimum.
Now, partially deriving ETC with respect to w; and y0 , we get

ETC

where

Step 5: Compute Q from (18) using k; y. If jQ  Q 0 j r ,

compute ETCQ ; k; y; n
and go to Step 6. Else set Q 0 Q and go to Step 3.
Step 6: If ETC n Z ETC, set
n
ETC n ETC; Q n Q ; yn y; k k; n n 1
n
and go to Step 2. Else n n  1 and stop.
The corresponding values of the control parameters for
nn n  1
give the optimal solution.

19

It is obvious from (14)(18) that the control parameters are not

independent of each other. So, to obtain a solution, we adapt the
iterative algorithm proposed by Ben-Daya and Hariga (2004). First,
the algorithm is initiated by setting y y0 where y0 is the original
percentage of defective items produced. Next, an initial value of Q
is calculated by setting the stochastic parameter of (18) equal to
zero. These initial values are used to determine the corresponding
value for k using (15). This is then used to update the values of y
and Q. This process is followed till a suitably stable solution is
reached. It is to be noted here that if the updated value of y is
found to be greater than the initial value y0 , then the updated
value is rejected. This follows intuitively since making an investment to improve process quality cannot end up making the
production process even more imperfect than it originally was.
Following the same argument, the value of y cannot be set less
than zero as well. The solution procedure can therefore be stated
as follows:
Algorithm 1.
Step 1: Set ETC n 1; n 1
p
Step 2: Set y y0 and compute Q 0 DGn=Hn; y0
Step 3: Compute k from (15) using Q 0 ; y and
R
k k1 z  kz dz
Step 4: Compute y from (17) using k; Q 0 . If y Z y0 , set y y0 .

21
22

It is obvious from (20)(22) that ETC increases with an increase

in warranty cost w and also with an increase in y0 , the original
percentage of defective items. This inference is intuitive as the
system's cost should increase if it has to bear a higher warranty
cost. Also, if the system produces items of very poor quality then it
makes sense to invest more to improve quality, thereby driving up
the total cost. An increase in implies that there is a greater
reduction in the number of defective items per dollar increase in
investment. That is, it costs less to improve the process quality and
hence ETC is a decreasing function of .
To further showcase the effects of the original process quality,
the investment option and other model parameters on the optimal
decisions, numerical studies are carried out in the following
section.

5. Numerical examples
For numerical studies, we consider the following data set:
D1000, P 3200, A 50, F35, K 400, L 10, hv 4, hb1 6,
hb2 10, s 0.25, x 2152, w 20, y 0.22, 100, 0:2,
0:0002
It is seen that for xed values of Q ; k, and n, the total cost
function ETC is convex in y0 o y r y0 . An illustration is provided
for y0 0:22 in Fig. 4. Further for given values of n and y, a 3Dgraph of ETC is shown in Fig. 5 which reects a convex surface.
It is obvious from Table 2 that with the increase in the warranty
cost w to be paid by the vendor, there is an increase in the optimal
total cost incurred by the supply chain. The optimal value of the
percentage of defective items also reduces with an increase in
warranty cost. This is intuitively correct since if the vendor has to
pay a higher warranty cost for producing defective items, it would
be benecial for him to reduce the number of defective items
produced. Similarly, with an increase in there is a decrease in the
total cost incurred and a corresponding decrease in the percentage
of defective items produced. This again makes sense from a
practical point of view since an increase in implies that there
is a greater reduction in the production of defective items per
dollar increase in investment. So, the production process quality is

O. Dey, B.C. Giri / Int. J. Production Economics 155 (2014) 222228

9000

Table 3
Effect of y0 .

ETC

8000

7000

6000

5000

227

0.00

0.05

0.10

0.15

0.20

y0

Qn

nn

yn

ETC n

Iyn

0.010
0.040
0.100
0.220
0.418
0.680

94.41
95.95
96.11
96.11
96.11
96.11

6
6
6
6
6
6

0.010
0.040
0.043
0.043
0.043
0.043

2833.26
3507.17
4420.81
5209.26
5851.12
6337.73

0.00
0.00
845.50
1635.95
2275.80
2762.38

Table 4
Effect of demand rate.

Fig. 4. The behaviour of ETC w.r.t. y.

d

Qn

nn

yn

ETC n

Iyn

800
900
100
1100
1200

84.67
90.46
96.11
91.87
96.97

6
6
6
7
7

0.052
0.047
0.043
0.039
0.037

4748.65
4989.17
5209.26
5409.21
5593.58

1438.67
1541.25
1633.95
1716.27
1793.26

Table 5
Effect of production rate.
P

Qn

nn

yn

ETC n

Iyn

2800
3000
3200
3400
3600

88.26
87.44
96.11
95.52
95.00

7
7
6
6
6

0.043
0.043
0.043
0.043
0.043

5168.81
5190.67
5209.26
5223.99
5237.00

1629.25
1630.66
1633.95
1634.88
1635.69

Fig. 5. The behaviour of ETC w.r.t. Q, k.

Table 2
Effect of parameters w and .
Parameter

Values

Qn

nn

yn

ETC n

20
24
30

96.11
95.77
95.43

6
6
6

0.043
0.037
0.030

5209.26
5374.27
5579.70

96.11
95.76
95.40

6
6
6

0.043
0.036
0.029

5209.26
4923.34
4606.14

0.00020
0.00024
0.00030

improved to a greater extent while the amount of investment

required for this improvement is reduced, thereby reducing the
optimal total cost.
An important insight that may be derived from Table 3 is that
the investment made to improve the production process quality is
not independent of the original quality. In other words, whether
any investment is required at all and, if required, to what extent it
would prove to be benecial depends on the original production
process quality. This is obvious from Table 3 which shows that for
very low values of y, there is no need for the vendor to invest in
improving the production process quality. However, as the value of
y increases, there is an increasing amount of investment required
on part of the vendor to optimize the supply chain.
It is also seen from Table 4 that an increase in the demand rate
causes the production lot size to increase and it requires an
increase in the amount of investment needed to optimize the
total cost. Table 5 illustrates the effect of the production rate P on
the system. Fig. 6 further illustrates the benet of making this

Fig. 6. The total cost incurred with and without additional investment.

investment by providing a comparison between the two scenarios

one where investment is not made and the other where the
vendor makes an investment.
Instead of using the logarithmic investment function, let us
now consider the same problem with a power investment function
of the form (Porteus, 1986):


Iy y   y0
where and are constants whose values can be estimated from
the past data.
Using the data given at the beginning of this section, and
74 000, 0:15, and adapting the suggested algorithm, the
optimal solution is obtained as follows: Q n 90:51, nn 7,
yn 0:1138 and the optimal cost incurred is ETC n 7281:86. It is

228

O. Dey, B.C. Giri / Int. J. Production Economics 155 (2014) 222228

noted that the results obtained for the power investment function
follow the same characteristics as of those obtained for the
logarithmic investment function.
6. Concluding remarks
In this paper, the problem of optimal investment of the vendor
in reducing the defect rate is analyzed for an integrated singlevendor single-buyer production-inventory system. It is assumed
that the production is imperfect and that, on receiving the items
from the vendor, the buyer inspects these items in a nonnegligible inspection period. It is further assumed that the vendor
makes an investment in improving the production process quality.
It is shown that this investment by the vendor reduces the
production yield rate of non-defective items and that, in case of
the vendor making such an investment, the integrated system is
better optimized in terms of minimizing the joint expected annual
total cost. Numerical studies also illustrate that an increased
demand rate requires an increased investment to minimize the
total cost. As a scope of future research, the model studied in this
paper can also be extended in terms of variable and/or controllable
lead-time and setup cost, inspection errors, variable shipment size,
etc. The model can also be extended for multiple buyers.
Acknowledgments
The authors are thankful to the anonymous referees for their
helpful comments and suggestions which have made signicant
improvement in the quality of the paper. The rst author gratefully
acknowledges the nancial support provided by NBHM, DAE, Govt.
of India under order No. 2/40(45)/2012-R&D-II/7842. The work is
also partially supported by the DST, Govt. of India under grant No.
SR/S4/MS: 565/09.
References
Banerjee, A., 1986. A joint economic-lot-size model for purchaser and vendor.
Decis. Sci. 17, 292311.
Ben-Daya, M., Hariga, M., 2000. Economic lot scheduling problem with imperfect
production process. J. Oper. Res. Soc. 51, 875881.

Ben-Daya, M., Hariga, M., 2004. Integrated single vendor single buyer model with
stochastic demand and variable lead-time. Int. J. Prod. Econ. 92, 7580.
Glock, C.H., 2009. A comment: Integrated single vendor-single buyer model with
stochastic demand and variable leadtime. Int. J. Prod. Econ. 122, 790792.
Glock, C.H., 2012. Lead timer eduction strategies in a single-vendor-single-buyer
integrated inventory model with lot size-dependent lead times and stochastic
demand. Int. J. Prod. Econ. 136, 3744.
Goyal, S.K., 1976. An integrated inventory model for a single supplier-single
customer problem. Int. J. Prod. Res. 15, 107111.
Goyal, S.K., 1988. A joint economic-lot-size model for purchaser and vendor: a
comment. Decis. Sci. 19, 236241.
Goyal, S.K., Crdenas-Barrn, L.E., 2002. Note on: economic production quantity
model for items with imperfect quality a practical approach. Int. J. Prod. Econ.
77, 8587.
Ha, D., Kim, S.L., 1997. Implementation of JIT purchasing: an integrated approach.
Prod. Plan. Control 8, 152157.
Harris, F.W., 1913. How many parts to make at once. Fact., Mag. Manag. 10, 135136
152.
Hill, R.M., 1997. The single-vendor single-buyer integrated production-inventory
model with a generalized policy. Eur. J. Oper. Res. 97, 493499.
Hill, R.M., 1999. The optimal production and shipment policy for the single-vendor
single-buyer integrated production-inventory problem. Int. J. Prod. Res. 37,
24632475.
Hsiao, Y.-C., 2008. Integrated logistic and inventory model for a two-stage supply
chain controlled by the reorder and shipping points with sharing information.
Int. J. Prod. Econ. 115, 229235.
Huang, C.K., 2004. An optimal policy for a single-vendor single-buyer integrated
production-inventory problem with process unreliability consideration. Int. J.
Prod. Econ. 91, 9198.
Lee, H.L., Rosenblatt, M.J., 1987. Simultaneous determination of production cycles
and inspection schedules in a production system. Manag. Sci. 33, 11251137.
Lin, H.J., 2013. An integrated supply chain inventory model. Yugosl. J. Oper. Res. 23,
87109.
Ouyang, L.Y., Wu, K.S., Ho, C.H., 2006. Analysis of optimal vendorbuyer integrated
inventory policy involving defective items. Int. J. Adv. Manuf. Technol. 29,
12321245.
Ouyang, L.Y., Wu, K.S., Ho, C.H., 2007. Integrated vendorbuyer inventory model
with quality improvement and lead-time reduction. Int. J. Prod. Econ. 108,
349358.
Pan, J.C.-H., Yang, J.S., 2002. A study of an integrated inventory with controllable
lead time. Int. J. Prod. Res. 40, 12631273.
Porteus, E.L., 1986. Optimal lot sizing, process quality improvement and setup cost
reduction. Oper. Res. 34, 137144.
Rosenblatt, M.J., Lee, H.L., 1986. Economic production cycles with imperfect
production process. IIE Trans. 18, 4855.
Salameh, M.K., Jaber, M.Y., 2000. Economic production quantity model with for
items with imperfect quality. Int. J. Prod. Econ. 64, 5964.
Shu, H., Zhou, X., 2014. An optimal policy for a single-vendor and a single-buyer
integrated system with setup cost reduction and process-quality improvement.
Int. J. Syst. Sci. 45, 12421252.
Scwaller, R.L., 1988. EOQ under inspection costs. Prod. Invent. Manag. 29, 2224.
Zhang, X., Gerchak, Y., 1990. Joint lot sizing and inspection policy in an EOQ model
with random yield. IIE Trans. 22, 4147.