See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/323082485

A critical spare part inventory control based on hazard function approach: A case
study in a garment company

Conference Paper  in  AIP Conference Proceedings · February 2018

DOI: 10.1063/1.5024081

CITATION READS

1 112

2 authors:

Intan Dewi Melinda Wakhid Ahmad Jauhari

Universitas Sebelas Maret Universitas Sebelas Maret
2 PUBLICATIONS   1 CITATION    105 PUBLICATIONS   225 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Undergraduate Final Assignment View project

Make or Buy Analysis and Quality Improvement Based on Tolerance Allocation View project

All content following this page was uploaded by Wakhid Ahmad Jauhari on 12 February 2018.

The user has requested enhancement of the downloaded file.

A critical spare part inventory control based on hazard function approach: A case
study in a garment company
Intan Dewi Melinda, and Wakhid Ahmad Jauhari

Citation: AIP Conference Proceedings 1931, 030022 (2018);

View online: https://doi.org/10.1063/1.5024081
View Table of Contents: http://aip.scitation.org/toc/apc/1931/1
Published by the American Institute of Physics
 A Critical Spare Part Inventory Control Based on Hazard
Function Approach: A Case Study in A Garment Company
Intan Dewi Melindaa) and Wakhid Ahmad Jauharib)

Production System Laboratory, Department of Industrial Engineering, Universitas Sebelas Maret, Surakarta,
Indonesia
a)
Corresponding author: [email protected]
b)
[email protected]

Abstract. Spare part procurement is a complex issue and requires an accurate analysis. Stock outs of spare part can leads
a great impact on production. Therefore, it is necessary to design the inventory control of spare parts that guarantee the
availability of spare parts needed for supporting the maintenance activity. This paper studies the inventory policy for
sewing machine spare part using hazard function to approximate the demand. Hazard function is the indicator of the
effect of ageing on the reliability of the system. It quantifies the risk of failure as the age of the system increases. We use
a continuous review policy based on Hadley Within Approach to calculate the optimum inventory level for critical spare
parts. There are four spare parts categorized as critical spare parts, which are needle plate, feed dog, rotary and binder
attachment. The optimal ordering quantity for needle plate, feed, rotary and binder attachment are 5 units, 17 units, 5
units, and 9 units, respectively and the reorder point are 2 units, 1 unit, 2 units and 1 unit, respectively. Finally, the
service level achieved by the proposed policy is in a range of 95.91%-97.93%, which indicates that the inventory level of
spare parts can be used to support the required parts in the maintenance activity.

INTRODUCTION
Spare parts management is essential to many industries, especially in a manufacturing company that uses
machines in their main activity such as a garment company. Excess inventory leads to high holding cost and a large
commitment of funds. On the other hand, stock outs can have a great impact on production. It is necessary to design
the inventory control of spare parts to improve productivity by reducing idle time and improve resource utilization.
Furthermore, spare part procurement is a complex issue and requires an accurate analysis of all the conditions and
factors that influence the selection of proper inventory policy. Most spare parts are characterized by an intermittent
demand, that is, occurs at certain period followed by long and several periods with zero demand. Intermittent
demands are particularly difficult to predict and shortage or excess may result in high costs [1].
Williams [2] was one of the earliest researchers to investigate inventory model for spare part with intermittent
demand and proposed a model similar to (Q,r) policy, considering a Gamma distribution demand. Popovic [3] and
Aronis et al. [4] used Bayesian model to forecast the spare part demand. Dekker et al. [5] developed a model with
different processing demands named as critical and non-critical with a Poisson distribution. Liao et al. [6] developed
a continuous review model with the assumption that the set of product grows at a constant rate and the interval
between failures follows a Weibull distribution. Supriyatna et al.[7] and Yunesa et al. [8] investigated inventory
level considering hazard function with Weibull Distribution. Aisyati et al. [9] determined an inventory policy for
aircraft consumable spare parts using a periodic review model.
For our literature review, an inventory model that predicts the demand by considering hazard function with
various probability distributions is rare. In this study, we overcome the problem of predicting the spare part demand
by analyzing the hazard function of each spare part. Hazard function calculation matches with the various
probability distributions, i.e. Exponential distribution, Weibull distribution, and Gamma distribution. Hazard

The 3rd International Conference on Industrial, Mechanical, Electrical, and Chemical Engineering
AIP Conf. Proc. 1931, 030022-1–030022-7; https://doi.org/10.1063/1.5024081
Published by AIP Publishing. 978-0-7354-1623-9/$30.00

030022-1
function is the indicator of the effect of ageing on the reliability of the system. It quantifies the risk of failure as the
age of the system increases.
In this study, we will determine which components is classified as a critical spare part and propose a proper
inventory model to control the inventories. First, we employ ABC classification to classify spare parts. Second, a
continuous inventory policy based on Hadley Within approach is used to determine the ordering quantity and
reorder point which minimizes the total cost. Third, we calculate the service level of the proposed policy to measure
its performance.

RESEARCH METHODOLOGY
We focus our study on a single needle sewing machine which is classified into the most critical machine in
production system due to its breakdown frequencies. Single needle sewing machine has 15 non-repairable
components. We categorize the spare part class using ABC classification method [10], [11]. We intend to focus our
study to determine the optimal inventory level for the spare part in class A. We conduct distribution test to
determine the probability distribution of time between failure (TBF) for each spare part. We use Easyfit 5.5 software
to test the distribution. Then, we define the hazard function based on its probability distribution using Wolfram
Mathematica 7.0. The functions defined by Wolfram Mathematica 7.0 are given as follows,

SDF[dist _, x _] : 1  CDF[dist , x] (1)

FM [dist _, x _] :  D[ SDF[dist , x], x] / SDF[dist , x] (2)

Where SDF is survival density function, CDF is cummulative density function and FM is force of mortality or
hazard function, while x is distribution parameter of each spare part.
The hazard functions used in this research are showed in equation (3), (4), and (5), respectively.

Hazard Function ( h(t ) ) for Exponental Distribution  λ (3)

α
β
t α 1e  βt
(α )
Hazard Function ( h(t ) ) for Gamma Distributi on  (4)
1 t α α 1  βt
1  β t e dt
(α ) 0
β 1
β t 
Hazard Function (h(t ) ) for Weibull Distribution  α (5)
α


Hazard function calculation that matches with the probability distributions, will be used to determine the
expected demand of each spare part by integrating the function. Expected demand from t1 to t2 (D) is showed in
equation (6) and expected demand during lead time (μ) is showed in equation (7).

t2
D   h(t ) dt (6)
t1

L t2
μ  h(t ) dt (7)
T t1

To determine the optimum inventory level for each critical spare part, we used a continuous review policy based
on Hadley-Within Approach with Poisson distributed demand.
The proposed algorithm to determine the optimum inventory level is as follows:
1. Calculate initial Q using equation (8).

030022-2
 2 AD
Q (8)
h

2. Find probability of shortage using equation (9).

 h xQ
 p( x)  1  (9)
x  r
1 Dxk

3. Define r using Cumulative Poisson Probability table with this following steps:
 Find the column headed by λ with the value of μ that calculated in previous section. λ is expected
number of occurrences or in this case, we define it with expected demand during lead time.
 Read the table and find values close to p(x)
 Find the rows headed by x that contain the values in step b. Value x represents reorder point.
4. Calculate expected of shortage by using equation (10).


E ( r )   ( x  r ) p ( x) (10)
x r
1

5. Find the new value of Q by using equation (11).


2 AD
(2kD x  ( x  r ) p ( x))
x  r
1
Q (11)
h

6. Repeat steps 2-3 until Q and r are in the same value.

In the final stage of this research, we calculate the service level to prove that the proposed policy can be used to
support the required parts in the maintenance activity. Service level for each spare part can be expressed in equation
(12).
1
SL  1   ( x  r ) p( x) (12)
Q

The following assumptions are used in our study:

1. The average demand D during lead time L follows a Poisson distribution.
2. Lead time is constant.
3. The order quantity is received all at once.
4. The cost of the ordering remains constant.
5. There is no limit order from supplier
The limitation of our study is that the optimal solution is calculated for each product or single product procurement.
So here, we consider a single inventory system.

RESULT AND DISCUSSION

Spare Part Classification

In the first stage, we categorize the spare part by employing ABC method. By following method, we can classify
which spare parts should be categorized as critical spare part. The manager should pay more attention to this kind of
spare part because it will give more impacts to the system performance. From the ABC classification, we found that
class A consists of 4 spare parts, class B consists of 4 spare parts and class C consists of 7 spare parts. The spare

030022-3
parts in class A give the contribution about 79.97%, spare parts in class B contributes 15.94% and spare parts in
class C contributes 4.09% to IDR volume. The results from ABC classification is given in table 1.

TABLE 1. Spare Part Classification

No Spare Parts Percentage Value Category

1 Needle Plate
2 Feed Dog
79.97% A
3 Rotary Hook
4 Binder Attachment
5 LED
6 Presser Foot
15.94% B
7 Knife
8 Needle Bar
9 Tension
10 Teflon
11 Bobbin
12 Plug 4.09% C
13 Limit Switch
14 Bobbin case
15 Bolt

Distribution Test
To determine the hazard function of each spare part, we need to determine the probability distributions and
parameters using EasyFit 5.0. The input parameter used to determine the probability distribution is Time Between
Failure (TBF) of each spare part. Time Between Failure is the length of time between consecutive failures [12]. The
distribution parameters resulted from the output of the software is given in Table 2.

TABLE 2. Spare part distribution test

Distribution Parameter
No Spare Part Distribution
λ α β
1 Needle Plate Gamma - 516.73 0.56128
2 Feed Dog Weibull - 37.991 0.62991
3 Rotary Hook Exponential 0.00184 - -
4 Binder Attachment Gamma - 507.63 0.51882

Determination of Hazard Function

After determining the probability distribution and parameters, we then calculate hazard function by following the
procedure explained in Section II. The hazard functions calculated using Wolfram Mathematica 7.0 are showed in
table 3.

030022-4
 TABLE 3. Spare part hazard function
No Spare Part Hazard Function (h(t))
0.018924 e 0.00193525 t
1 Needle Plate
t 0.43872 (1  Gamma Re gularized [0.56128,0,0.00193525t ]
0.0637121t 0.62991
2 Feed Dog
t 0.37009
0.00184e 0.t  0.00184e 0.00184t (1.
e 0.00184t )
3 Rotary Hook
1  e 0.00184t
Binder 0.0230893e 0.00196994
4
Attachment t 0.48118 (1  Gamma Re gularized [0.51882,0,0.00196994t ]

Determination of Expected Demand

The determination of expected demand is employed to predict how many spare part should be replaced from a
machine due to breakdown during a certain time. The expected demand can be calculated by integrating hazard
function with respect to t during 0 to 365 days (annual demand). The expected demand during lead time can be
calculated by applying the formula in equation (7). The results of hazard functon integration are shown in Table 4
and the results of D and μ are shown in Table 5.
TABLE 4. Integration of Hazard Function
No Spare Part Integration of Hazard Function
1 Needle Plate 1.Log[1.
0.630833 Gamma [0.56128,0,0.00193525 t ]]
2 Feed Dog 0.101145t 0.62991
3 Rotary Hook 0.00184t
Binder 1.Log[1.
0.584927 Gamma [0.51882,0,0.0016994 t ]]
4
Attachment

TABLE 5. Determination of D and μ

Rounding
No Spare Part Lead Time D365Days μ Rounding μ
D365Days

1 Needle Plate 18.68 19 0.36 1

2 Feed Dog 30.79 31 0.59 1
3 Rotary Hook 7 days
16.12 17 0.31 1
Binder
4
Attachment 19.19 20 0.37 1

Determination of Optimal Ordering Quantity and Reorder Point

In this research, we use a continuous review method based on Poisson distribution to calculate the optimal
inventory policy. We use an iterative procedure proposed by Hadley-Within to determine the optimal ordering
quantity and reorder point. The Poisson distribution is most useful for low volume demand or short lead time
situations. The input parameters of spare parts including spare part price, holding cost, ordering cost and shortage
cost are shown in table 6. Holding cost is formulated by 6% of spare part’s price, it is considering housing cost and
labor cost. In this research, shortage cost is formulated by 10 % of spare part’s price. The ordering cost is
determined by considering internet connenction cost as a major component in ordering activities.

030022-5
 TABLE 6. Spare Part Input Parameters
Spare Part Price Ordering Holding Cost Shortage
No Spare Part
(IDR) Cost (IDR) (IDR) Cost (IDR)

1 Needle Plate 189.000 11.340 18.900

2 Feed Dog 25.000 1.500 2.500
5.540
3 Rotary Hook 240.000 14.400 24.000
4 Binder Attachment 75.000 4.500 7.500

The optimal ordering quantity and reorder point are determined by applying an iterative procedure described in
section II. The optimal ordering quantity and reorder point for spare part are presented in table 7.
TABLE 7. Optimal Order Quantity and Reorder Point
No Spare Part Q (Unit) r (Unit)
1 Needle Plate 5 2
2 Feed Dog 17 1
3 Rotary Hook 5 2
4 Binder Attachment 9 1

Determination of Service Level

The service level is used to measure the performance of proposed inventory policy and it also represents the
expected probability of not hitting a stock out. The service level can be calculated by applying formula in equation
(12). The result given in table 8 shows that the service level is in a range of 95-98%, which indicates that the
inventory level of spare parts can be used to support the required parts in the maintenance activity.

TABLE 8. Service level

No Spare Part Service Level
1 Needle Plate 97.93%
2 Feed Dog 97.84%
3 Rotary Hook 97.93%
4 Binder Attachment 95.91%

CONCLUSION
In this research, we intend to determine the optimal inventory policy for critical spare parts using a continuous
review policy based on Poisson demand. An ABC method is used to categorize the spare parts based on their IDR
volume contribution. We consider the hazard function to predict the demand of spare part. The results from this
study show that the proposed method can achieved a high service level. It means that the proposed inventory policy
can guarantee the availability of spare parts needed for supporting the maintenance activity.

030022-6
 REFERENCES

1. Hua, Z. S., Zhang J.,Yang D. & S Tan. (2007). A New Approach of Forecasting Intermittent Demand For
Spare Parts Inventories in The Process Industries. Journal of the Operational Research Society, v. 58, p.
52-61.
2. Williams, T. M. (1982). Reorder Levels for Lumpy Demand. The Journal of the Operational Research Society,
v. 33-2,p. 185-189.
3. Popovic, J. B. (1987). Decision Making on Stock Levels in Cases of Uncertain Demand Rate. European
Journal of Operational Research. v. 32-2, p. 276-290.
4. Aronis, K. P., Magou I., Dekker R., & Tagaras G. (2004). Inventory Control of Spare Parts Using A Bayesian
Approach: A Case Study: European. Journal of Operational Research, v. 154, p. 730-739.
5. Dekker, R., Kleijn, M. J., & De Rooij, P. J. (1998). A Spare Parts Stocking Policy Based on Equipment
Criticality. International Journal of Production Economics, v. 5657,p.69-77.
6. Liao, H., Wang P., Jin T., & Repaka S. (2008). Spare Parts Management Considering New Sales. Annual
Reliability and Maintainability. Symposium, p. 502-507.
7. Supriyatna, Derry. N, Mustofa, H.Fifi & Zaini, Emsosfi. (2015). Sistem Persediaan Komponen pada Mesin
Cetak Berdasarkan Laju Kerusakan di PT. Karya Kita . Jurnal Online Institut Teknologi Nasional, 03, 252.
8. Yunesa, M. Randi, Adianto,H. & Susanty,S. (2016). Usulan Ukuran Pemesanan Optimal Suku Cadang Mesin
Grinding Berdasarkan Laju Kerusakan Menggunakan Metode Q (di Bengkel Pembuatan dan Service Turbin).
Jurnal Online Institut Teknologi Nasional, 04,38.
9. Aisyati, A., Jauhari, W.A & Rosyidi, C.N. (2014). Periodic Review Model for Determining Inventory Policy
for Aircraft Consumable Spare Parts. International Journal of Business Research and Managament. v.5-3.
10. Aisyati, A., Jauhari, W.A & Rosyidi, C.N. (2012). Determination Inventory Level for Aircraft Spare Parts
Using Continuous Review Model. International Journal of Business Research and Managament. v.4-1.
11. Chu, C.W., Liang, G.S. and Liao, C.T. (2008). Controlling inventory by combining ABC analysis and fuzzy
classification. Computers and Industrial Engineering. v.55
12. Dinesh Kumar, El-Haram, J.Knezevic, & U.Crocker. (2000). Reliability, Maintenance, and Logistics Support –
A Life Cycle Approach. Boston: Kluwer Academic Publisher.

030022-7
View publication stats