Beruflich Dokumente
Kultur Dokumente
M. Bonjakovia,*, M. Cobovib
a
b
Abstract
Spare part demand could significantly vary over a time. Even though there are periods without demand. Commonly used
lot sizing policies like Economic-Order
Order-Quantity, Lot-For-Lot
Lot and Period Order Quantity do not take these effects into
account. This research compares these policies with dynamic models, within which lot sizes are based on minimizing
total inventory cost. Appropriate example
xample is used to compare results within static and dynamic inventory models applied
to spare parts. Results show that the dynamic inventory models give the lower total inventory cost.
1. Introduction
Modern industry applications require the availability
and reliability of machines, which ensures, among other
things, the availability of spare parts and components at
the time of their needs. As the intensity of wear of
individual parts of the machine iss very different and
often unpredictable, it is necessary to stock a certain
amount of spare parts. However, ordering1 and
inventory holding2 costs are affecting performance. It is
therefore necessary to find the optimal order size that
will minimize total costs, while at the same time ensure
availability of spare parts at the time of their needs.
To find the optimal ordering plan,, there are different
mathematical models, but the question is which of them
give the best result in the issue of procuremen
procurement of spare
parts (HM. Wagner, 2004., R. Kleber, K. Inderfurth,
2009.).
In general, for solving this problem we can use
static and dynamic programming inventory models.
This is the sum of the fixed costs that are incurred each time a number of
spare parts is ordered. These costs are not associated with the quantity
ordered but primarily with the physical activities required to process the
order. These activities are: specifying the order, selecting a supplier, issuing
the order to the supplier, receiving the ordered
rdered goods, handling, checking,
storing and payment. It is also called setup cost.
2
Holding costs express the costs (direct or indirect) to keeping parts on
stock in a warehouse (warehouse space, refrigeration, insurance, etc
etc.
usually not related to the unit cost).
Inventory models
Static models
Dynamic models
Wagner-Within
2. Static Lot-Sizing
Sizing Models
2.1
The lead time is the time needed to get the spare part as indicated by the
supplier. It starts from the moment the supplier is informed until he
delivers the part on site.
Q* =
2.2
2 D Cn
h
(1)
TCt* = min(Ym,t )
(2)
(1 m t )
(3)
Boundary conditions:
Ordering is performed only when the inventory
level is zero,
Ordered quantity exactly corresponds to the
demands in observed time periods,
State of supplies x ordered quantity = zero
The following means that is never optimal to
order if there are any quantity on stock,
If it is optimal to order in the period m to satisfy
the demand for periods m to t, it is also optimal
to order in the period m for the periods (m, m+1,
., t).
Horizon theorem:
If it is in solving t periods optimal to order in the
period m to meet the demand in the period t, then in
resolving w periods (w>t) it is optimal to deliver order
in the period m or later:
If zt*=1 for the t period than zt*=1 for w periods
(w>t) and the ordering plan for the period t
remains unchanged (frozen)
If zt*=0 for t periods then zt*=0 or 1 for w
periods (w>t)
(4)
3.3
Where:
UCt - Average ordering and holding cost of
inventory per quantity unit.
P - The demand in period
b) Select the period in which t is UCt<UCt+1.
Which we denote as period t*.
c) Order the required quantity for the period 1 up
to t*.
d) Repeat the procedure for the period t=t*+1,
t*+2, t*+3, ,T
3.4
10 11 12 Total
5. Conclusion
Spare parts demand tends to be "lumpy," that is,
discontinuous and no uniform, with periods of zero
demand.
In general, dynamic models give better result than
static models for approximately 20%. The results of
dynamic methods depend on the value and mutual
respect of input data, and especially about the
relationship between the ordering and holding cost.
However, as it is evidently from the example and
additional analysis, the best result in determining the
optimal lot size of spare parts gives Wagner-Whitin
method.
Figure 7. Least period cost lot sizes
References
[1] HM. Wagner, Comments on Dynamic version of
the economic lot-size model. Management
Science, Vol. 50, No 12, December 2004, pp.
1775-1777
[2] S. Chand, A note on dynamic lot-sizing in a
rolling horizon environment. Decision Sciences,
Vol. 13, 1982, pp 113-119
[3] J.D. Blackburn, R. A. Millen, Heuristic lot-sizing
performance in a rolling schedule environment.
Decision Sciences, Vol.11, 1980, pp 691-701
[4] R. Kleber, K. Inderfurth, A Heuristic Approach for
Integrating Product Recovery into Post PLC Spare
Parts Procurement. Springer Berlin Heidelberg,
2009., ISBN 978-3-642-00141-3, pp. 209-214
[5] E. Silver, H. Meal, A heuristic for selecting lot
size requirements for the case of a deterministic
time varying demand rate and discrete
opportunities for replenishment. Production and
Inventory Management Journal, Vol. 14, No 2
1973., pp. 6474
[6] U. Wemmerlov, The part-period balancing
algorithm and its look ahead-look back feature: A
theoretical and experimental analysis of a single
stage lot-sizing procedure. Journal of Operations
Management, Vol. 4, No 1, 1983., pp. 2339
[7] James R. Evans, An Efficient Implementation of
the Wagner-Whitin Algorithm for Dynamic LotSizing. Journal of Operations Management, Vol. 5,
No. 2, , 1985., pp. 229-235