Professional Documents
Culture Documents
a r t i c l e in fo abstract
Available online 9 April 2010 Given a product design and a repair network for capital goods, a level of repair analysis determines for
Keywords: each component in the product (1) whether it should be discarded or repaired upon failure and (2) at
Maintenance which location in the repair network to do this. In this paper, we show how the problem can be
Supply chain management modelled as a minimum cost flow problem with side constraints. Advantages are that (1) solving our
Level of repair analysis model requires less computational effort than solving existing models and (2) we achieve a high model
Mixed integer programming flexibility, i.e., many practical extensions can be added. Furthermore, we analyse the added value of
Minimum cost flows modelling the exact structure of the repair network, instead of aggregating all data per echelon as is
common in the literature. We show that in some cases, cost savings of over 7% can be achieved. We also
show when it is sufficient to model the repair network by echelons only, which requires less input data.
& 2010 Elsevier B.V. All rights reserved.
1. Problem setting and literature If a component should be repaired, it should also be decided
where to do that. For example, if we consider military naval
For capital goods, such as military naval equipment, MRI-scanners, equipment, repairs can be performed on board the ship, at its
or trains, customers increasingly take total life cycle costs (LCC) marine base, a central depot, or even at the OEM. A network that
into account in their purchasing decisions (Ferrin and Plank, connects all ships, bases, depots, and the OEM is called a multi-
2002). Also, we see a trend that customers outsource activities for echelon repair network, see Fig. 2 for an example.
system upkeep to the original equipment manufacturer (OEM) In an early stage of the product life cycle, decisions upon
using service contracts that guarantee a certain service level discard or repair, and location of maintenance activities should be
against fixed yearly costs. For the OEM, this can be attractive, since taken. These decisions are covered by a level of repair analysis
selling services is generally more profitable than selling products (LORA). A LORA model should decide for a given product design and
(Deloitte, 2006; Murthy et al., 2004; Oliva and Kallenberg, 2003). repair network:
This means that it becomes important for the OEM to take the costs
of maintenance into account when designing new products. Costs which components to repair upon failure, and which to
of more reliable components can be earned back by lower discard,
maintenance costs during the product life cycle. for each component that will be repaired, where in the repair
Generally, capital goods are repaired by replacement, which network to do this, and,
means that a failed component is removed from the system and for each required resource (e.g. test or repair equipment),
replaced by a functioning one. A defective component can either where in the network to install it.
be discarded or repaired. If it is discarded, a new component
needs to be purchased. If the component is repaired, this often
means that a subcomponent failed and is replaced by a such that the lowest possible life cycle costs are achieved. Those
functioning one. The subcomponent should in turn be repaired costs consist of both costs that are variable in the number of
or discarded itself. The system is thus seen as a multi-indenture failures and fixed costs. Variable costs include costs of hiring
system such as shown in Fig. 1. service engineers and transportation of components; fixed costs
include costs for resources such as test equipment and tools.
The availability of the installed base is not considered in the
LORA. Instead, the spare parts stocking problem is solved using the
Corresponding author. Tel.: + 41 40 247 3947.
E-mail addresses: r.j.i.basten@tue.nl (R.J.I. Basten),
decisions that result from the LORA as an input (e.g., using METRIC-
m.c.vanderheijden@utwente.nl (M.C. van der Heijden), like methods, see Sherbrooke, 2004; Muckstadt, 2005). The goal in
j.m.j.schutten@utwente.nl (J.M.J. Schutten). this problem is to allocate inventory in a repair network such that
0925-5273/$ - see front matter & 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpe.2010.03.025
234 R.J.I. Basten et al. / Int. J. Production Economics 133 (2011) 233–242
Fig. 7. Source node. (Failures of component 1, shown by the number next to the
arc, originate at the source node.)
Fig. 9. Transformation node. (The flow going into the node results from a repair ‘r’
of component 1 at location 2, shown as ‘1,2,r’ in the node. Components 3 and 2
following example throughout this section: We have a two- flow out of the node, respectively.)
indenture system with three components (G1 ¼ f2,3g, see Fig. 5)
and a two-echelon repair network with three locations (see l. In this way, the variable costs of the LORA model are attached to
Fig. 6). We show an example of each node type, and after we have the arcs originating at the decision nodes; all other arcs have zero
introduced all the node types, we show the complete flow model. costs associated to them.
In all figures related to this example, a number next to an arc If the decision is taken to move component 1 from location 2 to
represents a component, whereas a letter next to an arc location 1, a decision has to be taken at location 1. This means that
represents an option (r¼repair, d ¼discard, m¼move). the arc representing the move option ‘m’ in our example (Fig. 8)
terminates at the decision node representing component 1 at
3.1.1. Source nodes location 1. In general, the arc representing the move decision for
Source nodes represent occurrences of failures of a certain component f at location k terminates at the decision node
subsystem at a certain system location. In our example only representing component f at location l, where k A Fl ðF1 ¼ f2,3gÞ.
failures in component 1 occur at locations 2 and 3. This means Note that at a node representing a component at the highest
that we have two source nodes in our flow model, of which the echelon, location 1 in our example, the ‘move’ option is not
one for location 2 is shown in Fig. 7. The flow out of this source available. The arcs representing the repair and discard options are
node is equal to the yearly number of failures of component 1 at discussed below.
location 2: m1,2. In general, for every component f A F s and every
location l A Ls , there is one source node vA V s with outflow
sv :¼ mfl . 3.1.3. Transformation nodes
Transformation nodes represent the repair of a parent
3.1.2. Decision nodes component. A flow representing failures in a parent component
If a subsystem at a system location fails, there are three goes into the node and is split according to the probability that a
options to choose from: failure in that parent component is due to each specific child
component. Each outgoing arc terminates at a decision node for a
child component. In our example, a failure in component 1 can be
move the component to the next higher echelon;
caused by a failure in component 2 or 3 ðG1 ¼ f2,3gÞ. The arc
repair the component, which means replacing a subcomponent
representing the repair decision ‘r’ for component 1 at location 2
or repairing the component directly;
(in Fig. 8) terminates at the transformation node that is shown in
discard the component.
Fig. 9. The two arcs originating at the transformation node
represent failures of components 3 and 2, respectively. Suppose
In the flow model, it means that an arc originating at a source that 50% of those failures are caused by failures in component 2
node terminates at a decision node v A V d .2 Every arc going out of and 40% are caused by failures in component 3, which means that
the decision node represents one of the available options. 10% of the failures is repaired directly ðq1,2 ¼ 0:5, q1,3 ¼
Fig. 8 shows the decision node for component 1 at location 2. 0:4, q1,1 ¼ 0:1Þ. If the outgoing arcs (v,w) and (v,u) represent
If arc (v,w) represents the repair option ‘r’, then the variable costs components 3 and 2, respectively, then gvw ¼0.4 and gvu ¼0.5.
for using this arc, cavw, are equal to the variable costs of option Since no further decisions need to be taken for the failures that are
‘repair’ for component 1 at location 2: cv1,2, repair. In general, the repaired directly, it is not necessary to model them.
variable costs, cavw, for using an arc ðv,wÞjvA V d are equal to cvfld, In general, transformation nodes vA V g represent the repair of
where arc (v,w) represents decision d for component f at location a parent component f by replacement of any of the subcompo-
nents g A Gf at location l. If arcs (v,w) represent components
2
The source node could also be integrated in the decision node, but for clarity g A Gf , then the factor of inflow in node v that flows out on arc
we prefer to have a distinction between source and decision nodes. (v,w), gvw, is equal to qfg.
R.J.I. Basten et al. / Int. J. Production Economics 133 (2011) 233–242 237
subject to
Xvw ¼ sv 8ðv,wÞ A Ajv A V s ð2Þ
X X
Xuv ¼ Xvw 8v A V d ð3Þ
Fig. 10. Sink node. (The flow that is discarded ‘d’ goes into the sink.) ujðu,vÞ A A wjðv,wÞ A A
Fig. 11. Example: network flow problema. aNote: A number next to an arc indicates the component that is represented by that arc. A letter indicates the decision that is
represented by that arc: r=repair, d=discard and m=move.
238 R.J.I. Basten et al. / Int. J. Production Economics 133 (2011) 233–242
X
Xvw ¼ gvw Xuv 8ðv,wÞ A Ajv A V g ð4Þ Table 1
ujðu,vÞ A A Input parameters.
Table 2
4. Computational experiments Comparison of optimization times (s).
instances of a product structure with corresponding failure rates component requires. In the first case, 70% of the components
and cost factors, see Appendix A for details. needs no resources, 20% needs one resource, and 10% needs 2
For our tests, we use various repair networks, of which some resources. To be short, we write ‘0.7–0.2–0.1’. The other case is
are balanced in terms of the number of system locations per ‘0.25–0.5–0.25’.
intermediate depot and some are not. We call them balanced and Summarizing, we have 7 experimental factors, namely the
unbalanced in the locations. The smallest balanced network number of intermediate depots, the number of system locations in
consists of one central depot, two intermediate depots and four the left half, the number of system locations in the right half, the
system locations (two per intermediate depot), see Fig. 2. We vary relative move costs in the right half, the relative repair costs in the
the number of intermediate depots (2 or 10) and the number of right half, the total number of resources, and the number of
system locations per intermediate depot (2 or 10). resources per component. This gives 2 2 2 3 3 2 2 ¼ 288
In the unbalanced networks, there are 2 system locations per combinations. For each combination, we generate 10 problem
intermediate depot for half of the intermediate depots and 10 instances as described in Appendix A. In this way we prevent that
system locations per intermediate depot for the other half of we draw conclusions based on one exceptional case.
the intermediate depots. Below, we call the left half of the
intermediate depots with the attached system locations the ‘left
half’. The other intermediate depot(s) with the attached system 4.2.2. Results
locations are called the ‘right half’. This holds for both balanced Table 3 presents for the 8 repair networks, the average time it
and unbalanced networks. takes to solve the exact and aggregated model. It also shows the
Besides being unbalanced in the locations, repair networks can average and maximal difference in optimal solution value for both
be unbalanced in terms of the costs. In our tests, costs in the left half models. In each test we vary over all other parameters, as
and at the central depot are always equal. Repair and move costs in explained in the previous section. We focus on the other
the right half can differ from the costs in the left half and at the parameters below.
central depot. We test what happens if the repair costs in the right We see that the aggregated models can be solved much faster
half are 0.5 or 2 times the repair costs in the left half (and at the than the exact models. Still, it took only 104 s to solve the most
central depot). We say that the relative repair costs are 0.5 or 2. In time consuming problem instance using the exact model. This
the same way, we test with relative move costs of 0.25 or 4. These means that the exact model can be solved fast enough to be used
values are chosen because for a European OEM, the costs of moving in practice.
components to Asia can be a number of times as high as those costs In networks that are balanced in the locations, the cost
in Europe. Repair costs can differ as well, but the relative difference reduction that can be achieved is limited: in our tests, it is 1.2%
is assumed to be smaller. Discard costs are assumed to be at maximum and only 8% of the problem instances leads to a cost
approximately the same at all locations, since a main part of these difference of more than 0.5%. This limited cost reduction means
costs are due to the costs of buying a new component. that it is hardly worthwhile to acquire the additional input data if
If we do not include spare parts costs in the LORA problem, the repair network is balanced in the locations, even if the
which is what happens usually in practice, we see that many network is unbalanced in the costs. However, in our tests, the
repairs are performed at a central location. The explanation is that spare parts costs that are added to the variable costs, are always
in that case only one resource of each type needs to be acquired. If balanced in the network, even if the remainder of the variable
repairs are performed decentrally, many resources are needed. In costs for repair and move are unbalanced. In a global repair
the problem instances that we discuss below, 86% of the costs for network, the spare parts costs may be unbalanced too, due to
resources would be made at the central depot if we do not include differences in lead times. If we vary the spare parts costs with the
spare parts costs. Integrating spare parts optimization into the move and repair costs in a network that is balanced in the
LORA is interesting future research, but it does not fit in the scope of locations, we see cost differences between the exact and
this paper. However, we add spare parts costs in a basic way, by aggregated model of over 5%. However, our basic way of
assuming lead times for all possible decisions (discard, repair and incorporating spare parts, does not allow us to analyse this in
move) and relating spare parts costs to these lead times. This is detail.
elaborated on in more detail in Appendix A. If the network is unbalanced in the locations, interesting cost
There are 10 or 25 resources (types of test equipment) and we reductions are achieved quite often: in our tests, the maximum
distinguish two cases for the number of resources that each cost difference is 7.4% and for 29% of the problem instances, a cost
Table 3
Comparison of various repair networks.a
Left half Right half Exact (s) Aggregated (s) Mean (%) Max (%)
a
It may seem strange that there is a difference between row 2 and 3 (2 intermediate depots and unbalanced network). However, these cases are not the same because,
for example, if the relative repair or move costs are not 1, this affects more system locations in row 2 than in row 3.
240 R.J.I. Basten et al. / Int. J. Production Economics 133 (2011) 233–242
Table 4
Comparison of three most important parameters.
Table 5
Example of different solutions.
reason is that repair at intermediate depot is much more In this paper we focussed on the LORA problem. The spare parts
expensive if there are ten intermediate depots, than if there are stocking problem is usually solved after solving a LORA problem.
only two, due to the number of required resources. As a result, We noticed that this sequential approach can result in performing
many more repairs are performed at the central depot instead of many repairs at a central location and discarding many compo-
at the intermediate depot, which Fig. 14 shows. Since many nents, which results in high spare parts costs. Therefore, it would
repairs are performed at the central depot, it is relatively be very useful to solve these two problems simultaneously. This
inexpensive to repair the subcomponents of those components might be done with an iterative approach using a LORA model and a
at the central depot too. Therefore, the discard costs in the spare parts stocking model, but one integrated model could be
network with ten intermediate depots are lower than in the used as well.
network with two intermediate depots. In general, we see that
the more system locations there are, the more attractive it
becomes to buy resources and repair components instead of Acknowledgments
discarding them. As a result, total costs increase not as much as
the number of system locations increases. The authors thank the anonymous referees for their comments,
We conclude that, using our test instances, modelling the which improved the quality of this paper. The authors gratefully
repair network exactly brings cost reductions of almost 2% on acknowledge the support of the Innovation-Oriented Research
average for networks that are unbalanced in the locations. In Programme ‘Integrated Product Creation and Realization’ (IOP IPCR)
some cases, the cost reductions are over 7%, which means that it is of the Netherlands Ministry of Economic Affairs and the support
worthwhile to model the repair network exactly for unbalanced of TRANSUMO.
networks. For networks that are balanced in the locations, cost
reductions are never higher than 1.2% in our test instances, which
means that it is doubtful whether the additional effort of Appendix A
acquiring all inputs is worth it. In this case, aggregating all data
does not lead to much higher costs. However, we do note that if In this appendix, we describe the random number generator
the costs for spare parts are not equal in the whole network, that we used in Section 4.2 to generate product structures, failure
due to a difference in lead time between different parts of the rates and cost factors. We give a summary of all parameter
global repair network, it might be necessary to explicitly model settings in Table 6.
the repair network. However, more research is needed on the In all tests, the system structure consists of 25 components at
integration of spare parts optimization into the LORA before this the first indenture level (subsystems), 125 at the second level and
question can be answered. 625 at the third level. Every second and third indenture
component is randomly attached to a lower indenture component
with equal probability per parent component. As a result, it may
5. Conclusions and directions for further research occur that a first indenture component has zero subcomponents
(which also occurs in practice, e.g., at Thales Nederland). The
We have modelled the LORA problem as a minimum cost flow annual demand per subsystem is drawn from a uniform distribu-
model with side constraints. This formulation allows us to model tion on the interval [0.01; 1]. The annual demand for a higher
all kinds of extensions in an elegant manner. Such extensions indenture component is a certain fraction of the demand for its
include repair probabilities, no-fault-found probabilities and parent component. The fraction of demand for the parent
equipment with a finite capacity. Besides that, we have shown component that is due to a failure in the child component is
that the LORA problem with all data aggregated per echelon level,
can be solved much faster using our formulation than it could Table 6
using existing formulations (a factor five on average for large Parameter settings.
problem instances).
Our model allows us to explicitly model the repair network Parameter Value(s) or range
instead of aggregating all information per echelon level, as is often # Components at indenture level 1 25
done in the literature. We have shown that with networks that are # Components at indenture level 2 125
unbalanced in the locations, substantial cost reductions can be # Components at indenture level 3 625
achieved by modelling the exact network (over 7% in our test
instances). If networks are balanced in the locations, then the Demand per subsystem [0.01; 1]
maximal costs reductions that can be achieved are much lower Fraction of demand for parent due to ½0:5=number of children;
(only 1.2% in our test instances), even if the network is child 1:25=number of children
Price of component [1,000; 100,000]
unbalanced in the costs. However, our research suggests that if
Repair costs (fraction net price) [0.1; 0.4]
lead times differ across the repair network, significant cost Discard costs (fraction gross price) [0.75; 1.25]
reductions may be achieved by modelling the exact network, Move costs (fraction gross price) 0.01
even for networks that are balanced in the locations. To be able to
analyse this, integration of spare parts optimization in the LORA is Resource costs [10,000; 1,000,000]
necessary. # Resources 10 & 25
We have shown that in some cases, only small cost differences Fraction of components requiring 0–1–2 0.7–0.2–0.1 & 0.25–0.5–0.25
exist between using the exact and the aggregated network, resources
Repair lead time at echelon level 1 1 month
although the decisions that are taken differ a lot. If there are
Repair lead time at echelon level 2 1 month
multiple ways to achieve almost the same total costs, there can be Repair lead time at echelon level 3 3 months
other, more qualitative reasons to choose for another solution Move lead time between two echelons 0.5 months
than the one with the lowest total costs. Future research could (1 & 2, 2 & 3)
lead to an approach that results in multiple alternative solutions Discard lead time 6 months
Safety factor 2
that differ a lot in terms of decisions, but lead to almost the same Holding costs 0.3
optimal solution value.
242 R.J.I. Basten et al. / Int. J. Production Economics 133 (2011) 233–242
drawn from a uniform distribution on the interval are 70%, 20% and 10%, respectively. To be short, we write ‘0.7–0.2–
½0:5=number of children; 1:25=number of children, with a max- 0.1’. We also perform tests with settings ‘0.25–0.5–0.25’.
imum of 1. Summing the fractions of all children leads to a value We also need to take spare part costs into account, see
between 0.5 and 1.25. If this summation is lower than 1, then a Sections 1 and 4.2.1. As described in those sections, modelling
certain fraction of the failures in the parent can be repaired spare parts costs is not straightforward and should be item-based
directly, without replacement of a child component. If the since a system approach, such as the VARI-METRIC type approaches
summation is higher than 1, then a certain fraction of the failures (see, for example, Sherbrooke, 2004), is not possible. Our
in the parent leads to the replacement of more than one child approach is based on the difference in lead times that result from
component. different LORA decisions. We assume that repair at the system
For each component, we draw a net price, excluding the costs location or at the intermediate depot takes a month, whereas
of the children, from a shifted exponential distribution with shift repair at the central depot takes three months. Discard (and
factor 1,000 and rate parameter l ¼ 7=ð100,0001,000Þ. In this buying a new item) has a lead time of half a year. Moving a
way, on average 1% of the components get a price higher than component to a higher echelon leads to an additional lead time of
h100,000. We draw a new price for these components to avoid half a month. Using these lead times and a safety factor of 2, we
odd problem instances. The reason for our choice is that most estimate spare parts inventory as being the safety factor times the
systems have a large diversity of items in price, but there are lead time demand. To compute the corresponding holding costs,
considerably more cheap items than expensive items. The we use an inventory carrying charge of 30%. We tested other
cheapest items (in our case items with a price below h1,000Þ approaches and safety factors, but found that this approach
are usually omitted from a regular LORA, because they are generally leads to reasonable results.
discarded by default.
Using these prices, we calculate the variable costs as follows: References
Ahuja, R.K., Magnanti, T.L., Orlin, J.B., 1993. Network Flows. Theory, Algorithms,
repair costs as a fraction of the net component price are drawn
and Applications. Prentice-Hall, Englewood Cliffs (NJ).
from a uniform distribution on the interval [0.1; 0.4]; Alfredsson, P., 1997. Optimization of multi-echelon repairable item inventory
we recursively add the price of each child to its parent to get systems with simultaneous location of repair facilities. European Journal of
the gross item price. We do this after calculating the repair Operational Research 99, 584–595.
Barros, L.L., 1998. The optimization of repair decisions using life-cycle cost
price, since repair of a parent means replacement of the child parameters. IMA Journal of Mathematics Applied in Business & Industry 9,
that was defective and taking a decision for the child, thus 403–413.
incurring costs for the child. Discarding or moving a parent Barros, L.L., Riley, M., 2001. A combinatorial approach to level of repair analysis.
European Journal of Operational Research 129 (2), 242–251.
does not lead to a decision for the child components, and thus Basten, R.J.I., Schutten, J.M.J., Van der Heijden, M.C., 2009. An efficient model
not to incurring costs for the child components either; formulation for level of repair analysis. Annals of Operations Research 172 (1),
the discard costs as a fraction of the gross item price, including 119–142.
Basten, R.J.I., Van der Heijden, M.C., Schutten, J.M.J., 2010. Practical extensions to
children, are drawn from a uniform distribution on the interval the level of repair analysis. BETA working paper 304, submitted for publication.
[0.75; 1.25]. The 100% would be just the costs of replacing a Brick, E.S., Uchoa, E., 2009. A facility location and installation or resources model
defective component by a new one. However, there may be for level of repair analysis. European Journal of Operational Research 192 (2),
479–486.
additional disposal costs to satisfy legal requirements for
Deloitte, 2006. The service revolution in global manufacturing industries.
waste processing. On the other hand, some parts of a defective EDCAS, 2009. /http://www.tfdg.com/edcas.phpS, last checked on 9 May 2009.
component may be recycled or re-used. Therefore, we vary the Ferrin, B.G., Plank, R.E., 2002. Total cost of ownership models: an exploratory
study. Journal of Supply Chain Management 38 (3), 18–29.
disposal costs around the price of a new component;
Muckstadt, J.A., 2005. Analysis and Algorithms for Service Parts Supply Chains.
the move costs are always 1% of the gross item price. Springer, New York (NY).
Murthy, D.N.P., Solem, O., Roren, T., 2004. Product warranty logistics: issues and
challenges. European Journal of Operational Research 156, 110–126.
We perform tests with 10 and 25 resources. The price of each Oliva, R., Kallenberg, R., 2003. Managing the transition from products to services.
International Journal of Service Industry Management 14 (2), 160–172.
resource is drawn from a shifted exponential distribution with
PRICE HL, 2007. /http://www.pricesystems.com/products/price_hl.aspS, last
shift factor 10,000 and rate parameter l ¼ 7=ð1,000,00010,000Þ checked on 25 July 2007.
in the same way as described above for the prices of the Saranga, H., Dinesh Kumar, U., 2006. Optimization of aircraft maintenance/support
components. As a result, prices vary between h10,000 and infrastructure using genetic algorithms—level of repair analysis. Annals of
Operations Research 143 (1), 91–106.
h1,000,000. We randomly assign components to resources, such Sherbrooke, C.C., 2004. Optimal Inventory Modelling of Systems. Multi-echelon
that the percentage of components that needs 0, 1 and 2 resources Techniques, second ed. Kluwer, Dordrecht, The Netherlands.