make to stock or make to order: the decoupling point in the food processing industries

*Corresponding author. Tel.: #31-50-3637345; fax: #31-50-3632032.

E-mail address: [email protected] (D.P. van Donk).

Int. J. Production Economics 69 (2001) 297}306

Make to stock or make to order:The decoupling point in the food processing industries

Dirk Pieter van Donk*

Faculty of Management and Organization, University of Groningen, P.O. Box 800, 9700 AV Groningen, Netherlands

Received 13 April 1999; accepted 29 February 2000

Abstract

Food processing industries experience growing logistical demands, growing variety in products and intense competi-tion. As a reaction companies try to produce more on order. Often managers "nd it di$cult to decide which products tomake to order and which products to stock. This paper develops a frame that is an aid for managers in balancing thefactors and characteristics of market and production process that in#uence such decisions. The frame is based on thegeneral decoupling point concept by Hoekstra and Romme, which is adapted to the speci"c characteristics of the foodprocessing industry. Its usefulness is illustrated in a case study. Some directions for further research are given. ( 2001Elsevier Science B.V. All rights reserved.

Keywords: Decoupling point; Market response; Make-to-order; Food processing

1. Introduction

For many years it was a common policy for foodprocessing companies to produce in large batchesto keep production costs low and limit the numberof set-ups. This seemed to be a good policy. The lastdecade showed a number of changes, graduallygrowing in signi"cance. The background andcauses of these changes can be summarised underthree main themes.

Firstly, consumers' wishes seem to changein an ever growing rate, causing an increase in

packaging sizes, the number of products as wellas in the number of new products introduced,e.g. [1].

Secondly, many retailers are restructuring theirsupply chain both in a physical and information#ow sense. The aims are reduction in inventories,faster replenishment and shortening of cycle times.The result for food processing industries is thatlogistical performance needs to be improved: fasterand more dependable. There are some examples ofreductions in lead-time from 120 hours in the pastto 48 hours now and still further reductions are tobe expected.

Thirdly, the above-mentioned changes have tobe realised in a market which can be characterisedby low margins in retailing and mergers andacquisitions in retail chains [1]. Both lead toa downward pressure on prices paid to producers.

0925-5273/01/$ - see front matter ( 2001 Elsevier Science B.V. All rights reserved.PII: S 0 9 2 5 - 5 2 7 3 ( 0 0 ) 0 0 0 3 5 - 9

In summary, food processing industries have todeliver a greater variety of products, have to meethigher logistical demands, while keeping costs aslow as possible. These demands are especiallyvisible in those industries that produce for theconsumer-market and their direct suppliers.

This article explores more #exible production asa possibility to cope with these market demands. Assaid above, the industry standard is to produce tostock. Producing more #exible requires that (partof the) customer orders have a direct impact onproduction orders. However, usually there are cer-tain limits and not all products for all customerscan be made to order. Therefore an importantquestion is which products, product families orproduct}market combinations can and should beproduced to stock and which can be made to order.To answer this question the general frame ofHoekstra and Romme [2] is used. Especially theirconcept of decoupling point, which is de"ned as `thepoint that indicates how deeply the customer orderpenetrates into the goods #owa. Hoekstra andRomme distinguish a frame in which a number ofproduct and market characteristics, and processand stock characteristics in#uence the location ofthe decoupling point. This frame will be adapted tothe speci"c characteristics of the food processingindustry.

The main question to be answered in this articleis how managers in food processing industry can besupported in deciding if some of their products canand should be produced to stock or be made toorder and, if there are some, which should beselected. In investigating this, we also add to ourknowledge of the usefulness of the decoupling pointconcept in this particular type of industry. So far,little attention has been paid to this type of ques-tions in the literature.

The article is organised as follows. First, we willintroduce the general frame of the decouplingpoint. Next, the speci"c characteristics of food pro-cessing industries are presented. Then, the "ndingsof these two sections are combined to arrive ata concept for the decoupling point tailored to thefood processing industry. Section 4 will show theusefulness of the concept in a case study. Lastly,conclusions are drawn and some remarks regardingfurther research are made.

2. The decoupling point concept

The background of the concept of decouplingpoint lies in the observation that within productionmanagement and logistical management attentionhas been paid to all kind of separate elements ina production chain, without notice for the need foran integrated framework. In developing the frameof the decoupling point (DP) Hoekstra and Romme[2] intend to furnish a concept for integral control.Integral control in this context means planning andmanagement of the goods #ow from purchasedmaterials to delivery takes place, based on the char-acteristics of the product}market combination,within a suited organisational and control structure(also see, [3]).

An important concern in designing integral con-trol is "nding a balance in the costs of procurement,production, distribution and storage against thecustomer service to be o!ered.

The result of this balancing establishes thedecoupling point for a certain product}marketcombination. The decoupling point separates thepart of the organisation oriented towards activitiesfor customer orders from the part of the organisa-tion based on forecasting and planning [2, p. 6], or,in other words the decoupling point is the pointthat indicates how deeply the customer order pen-etrates into the goods #ow [2, p. 66].

The decoupling point is important for a numberof reasons:

f It separates the order-driven activities from theforecast-driven activities. This is not only impor-tant for the distinction of di!erent types of activ-ities, but also for the related information #owsand the way the goods #ow is planned and con-trolled.

f It is the main stock point from which deliveriesto customers are made and the amount of stockshould be su$cient to satisfy demand in a certainperiod.

f The upstream activities can be optimised in someway, as they are based on forecasts and are moreor less independent from irregular demands inthe market.

According to [2] "ve possible DP positions cor-respond to "ve basic logistical structures: make and

298 D.P. van Donk / Int. J. Production Economics 69 (2001) 297}306

Table 1Determinants of the decoupling point

Product and market characteristics Process and stock characteristics

Required delivery reliability Lead times and costs of steps in the (primary) processRequired delivery time Controllability of manufacturing and procurementPredictability of demand Costs of stock-holding and value added between stock pointsSpeci"city of demand Risk of obsolescence

Fig. 1. Business characteristics classi"ed according to their nature and in#uence on the DP [[2], p. 71].

ship to (local) stock; make to (central) stock; as-semble to order; make to order; purchase and maketo order. As stated the determination of the posi-tion of the DP depends, in general on two sets ofcharacteristics: product and market characteristics,and process and stock characteristics, which aresummarised in Table 1.

Their respective in#uence is depicted in Fig. 1 [2,p. 71]. Fig. 1 should be interpreted as follows. Foreach of the terms mentioned the in#uence on thelocation of the DP is shown. e.g. irregular marketdemand will (if all other things remain stable) havean upstream e!ect on the location of the DP, whileshort delivery times will force the DP more down-wards, towards the make-to-stock position.

The DP concept is a valuable tool in describingand analysing production processes and the goods#ow of organisations. However, from the existingliterature it is hard to derive rules for locating orchanging the position of the DP or procedures forbalancing the relevant characteristics. Also, thee!ects of having more DP's (for di!erent prod-uct}market combinations) in one factory and theconsequences for planning have been largelyneglected. Recently, [4] combines the location ofthe DP and the location of the capacity constraintinto one matrix to derive the characteristics of anintegrated material and capacity-based masterschedule. Further work into this direction seemsa good addition to the DP concept.

D.P. van Donk / Int. J. Production Economics 69 (2001) 297}306 299

Fig. 2. An example of a production process in food processing.

3. Characteristics of food processing industries

This section introduces the food processingindustries from a production management pointof view. For the purpose of this paper it is usefulto distinguish between companies that produce in-termediate products from natural materials andcompanies that process these intermediate prod-ucts further into consumer or industrial products.The "rst category involves mills, abattoirs, sugarre"neries, etc. while the second category entailsproducers of canned meat, bakeries, producers ofchocolates, etc. Of course, this distinction is roughand some of the producers of intermediate prod-ucts, produce some consumer products as well (e.g.a sugar re"nery producing sugar cubes). We willpay attention to the second category. An exampleof a food processing process can be depicted as inFig. 2.

Instead of packaging, other processes mightoccur and sometimes products are pasteurisedor sterilised before packaging. In many cases, stockpoints as depicted can only store temporarily, dueto the perishability of intermediate products orbecause the capacity for storage on the shop#oor islimited.

From the literature [5}9] the following enumer-ation of characteristics of food processing industrycan be compiled [10].(1) Plant characteristics

(a) Expensive and single-purpose capacitycoupled with small product variety andhigh volumes. Usually, the factory showsa #ow shop oriented design.

(b) There are long (sequence-dependent) set-uptimes between di!erent product types.

(2) Product characteristics(a) The nature and source of raw material in

food processing industry often impliesa variable supply, quality, and price due tounstable yield of farmers.

(b) In contrast with discrete manufacturing,volume or weights are used.

(c) Raw material, semi-manufactured prod-ucts, and end products are perishable.

(3) Production process characteristics(a) Processes have a variable yield and process-

ing time.(b) At least one of the processes deals with

homogeneous products.(c) The processing stages are not labor inten-

sive.(d) Production rate is mainly determined by

capacity.(e) Food industries have a divergent product

structure, especially in the packaging stage.(f) Factories that produce consumer goods can

have an extensive, labor-intensive packag-ing phase.

(g) Due to uncertainty in pricing, quality, andsupply of raw material, several recipes areavailable for a product.

In most cases a limited number of these character-istics is present. We use the list above for examin-ing, describing and analysing real-life situations.Each of the factors presented, has to be taken intoaccount for planning and scheduling purposes. E.g.high set-ups and an orientation to use capacity asmuch as possible, cause planning of long produc-tion runs and stocks of end products.

4. Decoupling point in food processing industry

The previous sections elaborated upon the gen-eral in#uences of market and process character-istics on the position of the decoupling point andon the characteristics of the food processing indus-try. This section will relate the two. For each factorthe e!ect on the decoupling point for a certainproduct/market combination will be discussedunder a ceteris paribus clause for other factors. This


Table 2Market characteristics and their in#uence on the decoupling point

Market characteristic Presence/value in food E!ect on DP

Delivery reliability High DownstreamDelivery time Short DownstreamPredictability of demand (rather) unpredictable Upstream through information sharingSpeci"city of demand Great variety end products (with common recipes) Upstream possibilities

analysis is performed from the point of view ofa producer of food products, delivering to retailers(nowadays, usually supermarket chains, or integ-rated combinations of wholesaler with a number ofoutlets).

With respect to the market characteristics we mayconclude that, as was already stated in the intro-duction, delivery times are usually short and cus-tomers (such as retail chains) need a high reliability.Both tend to have a downstream e!ect on thedecoupling point. These short delivery times seemto go hand in hand with a relative unpredictabilityof the demand from the point of view of theproducer. In a number of cases the retailers try topass all the uncertainty in demand onto the pro-ducers by asking instant delivery within very shorttime, without adequate support in forecastingdemand. Nowadays better opportunities emergein forecasting by analysing point-of-sale and scan-ning data. Better co-operation and sharing of thesedata between retailer and producer could bringmutual bene"ts within easy reach and even openpossibilities for a longer lead time. Although actualdemand of the consumers might be still erratic,such joint e!orts in forecasting might improveoverall performance of the supply chain. A lastpoint to pay attention to, is the speci"city of de-mand. As noticed above, food processing industrieshave divergent product structures. Often the diver-sity of products originates from the large number ofpackaging sizes, labels and brands. Speci"city alsocomes into being through the best-before-date. It isinteresting to note that many products have a tech-nical best-before-date which is reasonably long.However, retailers do not accept succeeding delive-ries with identical best-before-dates. The result isthat from a technical point of view products are still

fresh, but from a commercial point of view obsoleteand in fact, extremely perishable. All in all, thefactor speci"city o!ers some potential for an up-stream repositioning of the decoupling point ifcommonality can be used and intermediate storagebetween processing and packaging is technicallypossible. The above discussion is summarisedin Table 2.

With respect to the process and stock character-istics it is clear that part of food processing indus-tries have an uncontrolled process with variableyield and, due to variability in natural materials,variable quality of products. Such factors causea downstream e!ect on the decoupling point,because storing a product after the uncontrolledprocess safeguards undisturbed delivery. Cleaningtimes and set-ups (which are often sequence depen-dent) are an important factor in production in foodprocessing industry. If these are large, the e!ect onthe decoupling point will also be downstream. Anupstream e!ect on the decoupling point might beexpected for the remaining factors: stock levels (andcosts) and risk of obsolescence. Both factors relateto the nature of food: it's perishability. Retailersalways want the most recent `best-beforea andstock might easily become out-of-date due to thebest-before on the product. This might occur even ifthe technical shelf-live is still quite long. The valueof stock is related to this aspect. There are twomajor factors that contribute to the value: the valueadded in production and the value of the materialspurchased. A high value added in production willhave an upstream e!ect on the decoupling point,as it is "nancially bene"cial to store low-valuegoods instead of higher valued end products. Thevalue of the materials is only a relevant factor ifpurchasing can be postponed until actual usage in


Table 3Process characteristics and their in#uence on the decouplingpoint

Process characteristic Presence/value in food E!ect on DP

Lead times & costs Relevant set-up times DownstreamControllability (sometimes) low DownstreamValue added and costs

of stock-holdingUnclear (in general) *

Risk of obsolescence High Upstream

production. In a number of food processing indus-tries e.g. co!ee, co!ee beans of a desired quality canonly be purchased once a year. In that case there isno "nancial reward for changing the decouplingpoint. In factories for canned meat, however, thepurchasing of meat can be postponed until produc-tion starts. In such a case the aggregate level ofinventories can be lower and an upstream move-ment of the decoupling point may have a "nanciale!ect. The above discussion is summarised inTable 3.

There are some special points of interest in thefood processing industries, which may in#uence thepossible location of the decoupling point.

The "rst point relates to the capacity of theproduction system. In many cases there is somekind of a bottleneck capacity: either in the process-ing stage or in the packaging stage. In generala bottleneck capacity is less suited for producing toorder, as more variation can be expected in directcustomer orders then in stock orders. Conse-quently, if the packaging department has limitedcapacity the possibility for locating the decouplingpoint upstream from the packaging stage will belimited. If the processing stage has limited capacitythen locating the decoupling point betweenprocessing and packaging is possible, setting asidepotentially restricting characteristics of products.

The second point is related to regulations andlaws aiming at protecting the safety of consumers.More and more, food processing companies haveto obey to strict regulations, which demand demon-strable safety through HACCP (hazard analysis ofcritical control points) and traceability. This putsnot only a pressure on registration and information

#ows, but might also restrict the possible locationof the decoupling point as intermediate inventoriesare less controllable and traceable than inventoriesof materials and end products.

A third point of interest is the attention given toand the introduction of the concept of ECR (e$-cient consumer response). One of the important socalled improvement concepts within this frame iscontinuous replenishment. For food processing in-dustries this concept can result in two kindsof consequences. On the one hand, retailers candemand short delivery times, without communicat-ing forecasts or information on sales. This results ina downstream e!ect on the decoupling point, asmentioned before. On the other hand, this mightresult in a closer cooperation between retailer andproducer and more intensive exchange of relevantinformation on sales will take place. As a resultmore production to order is possible.

A fourth point, which is interesting for the loca-tion of the decoupling point, makes a small adapta-tion to the assumptions and statements made in theintroduction. In food processing industries produc-tion to stock is usual, but especially in exporting"rms we observe a combination of make-to-stockand make-to-order. Usually the (large) orders forforeign countries are separately dealt with and pro-duction to order is possible due to longer leadtimes. Often production to order is necessary asthese orders arrive with great intervals and speci-"cation and magnitude is not known in advance.For our discussion this observation is importantbecause these companies already cope with di!er-ent decoupling points and combine make-to-stockwith make-to-order.

5. An illustrative case

The case concerns a manufacturer that buildsa new facility. In the old facility production to stockwas the only possibility due to the technologicallimitations. The new facility has the possibility tostore a number of semi-"nished products in silosand the management of the factory aims at produ-cing as much as possible to order. So, the questionwas which products were to be made to stock,which products to semi-"nished product and which


Fig. 3. The production process of the case.

products totally to order. In other words, the ques-tion is where the decoupling point ought to belocated for each product. To answer this questionthe afore-mentioned characteristics were analysed.We will brie#y describe the results.

5.1. Process

The production process has three steps: process-ing, granulation and packaging (see Fig. 3).

The processing stage consists of several steps:mixing of raw materials according to recipe anda number of subsequent processing steps which areexecuted without interruption or intermediate stor-age. Then the semi-manufactured product may bestored in one of the silos or can be granulated,directly. Next, the product is separated into severalfractions that are made of the same recipe but di!erin the size of the granule. In the last step theproduct is put into big bags or smaller bags of5}25 kg with (sometimes) a client-speci"c text on it.Now, the "nished product can either be delivered tothe customer or can be stored. Throughput timesfor each batch are approximately two hours forprocessing and half an hour for granulation. Theproduction rate (in kilogram/hour) of the "rst stepis about half the rate of the second step. Set-up-times are relatively large for both processes andare sequence dependent. The production process isreliable in quality and amount of output.

5.2. Market

The company produces some 200 di!erent prod-ucts, which di!er in recipe (40 di!erent recipes),granule (30 di!erent sizes) and packaging. Demandis stable in an aggregate way but irregular (both inamount and time) and not easy to forecast on

a detailed day-to-day or even week level. Fiverecipes (which are the basis of several "nished prod-ucts) account for about 70% of total demand. Thenumber of customers is high and even the largestcustomers have a share in total volume of less than10%. The lead time for delivery has a standard of5 days, but quite a few customers ask for shorterdelivery. On the other hand, some important cus-tomers order in a regular way with a lead time of2 weeks. Some customers ask the company to keepa certain amount of stock dedicated to them, forimmediate delivery. Customers ask for dependentdelivery, as is usual nowadays.

5.3. Product and stock

The product can be kept for almost half a year.However, the producer has to guarantee his cus-tomers a shelf life of 4 months at least. That meansthat slow-moving products have a risk of becomingobsolete. As said before the management of thecompany aims at lower inventories.

Each of the three storage points has a limitedcapacity: the capacity of the silos is most restrictinghaving a storage capacity of a little more then1 week of average sales.

Due to the nature of the product, the qualityrequired and the way it is produced, there arerestrictions with respect to the minimal batch sizesin the processing stage, the granulation andpackaging stage. These restrictions are incorpor-ated in the production technology and will betreated as limitations that cannot be changed.

5.4. Locating the decoupling point

From the description of the production process itis clear that, theoretically, three possible decoup-


ling points exist: the stock of raw materials, thestock of semi-"nished products (in silos) and thestock of "nished products. The storage space forsemi-"nished products is limited by the numberand capacity of the silos. So, in this case it seemsnatural to investigate "rst, which products shouldbe stored there.

Firstly, we observed that a number of semi-"n-ished products is used in one "nished product,only. That kind of products can be excluded fromstorage in the silos. The underlying logic is thatthese products are client speci"c as soon as the "rstoperation is started. Consequently, intermediatestorage has no advantage and production can beeither to stock or to order, depending on theamount asked for and the lead time needed. Aboutone-third of the semi-"nished products is soexcluded.

A second observation is related to the limitedcapacity in the processing stage. To use capacity ase$ciently as possible the number of set-ups andcleaning times in the processing stage has to berestricted. Thus, it is advantageous to produce inrelatively large bat ches, which is only necessary forproducts having a relatively large aggregate de-mand. The products with su$ciently large demandare subsequently stored in the silos to limit thestock of "nished products and still being able todeliver fast. In applying this rule, we could choosethe right number of products to be stored in thesilos. For the products stored in the silos the dilem-ma between e$ciency and responsiveness seems tobe solved: processing is in large and e$cientbatches, while granulation and packaging isconducted in response to the market.

Some products that we would like to produce toorder (e.g. due to irregular demand) will be produc-ed to stock if the amount asked for is smaller thanthe minimum batch sizes. The minimum batch sizecauses an amount of stock of the "nished product.This stock has a relatively high risk of becomingobsolete. Resolving this problem is not straightfor-ward and a!ects the market strategy. For productswhich are ordered irregularly, but in largeramounts than the minimum batch size there is noproblem and these products are of course producedto order. In other words, the decoupling point is thestock of raw materials.

5.5. Results

As a result of the above location decisions about75% of the number of the "nished products will beproduced on order. An important reason to havestock of the other articles is the very short lead timeasked for (often in combination with the wish ofcustomers to let the producer keep an amount ofstock dedicated to them). Another reason is regu-larity in demand: each week a number of orders fora certain product.

Changing the DP for a number of productsa!ects other performance measures as well.However, it is hard to make a comparison with thepreviously existing situation because a completenew factory is built. Still we can highlight some.First, it is important to note that customer service(in terms of dependability and speed) is improved,largely due to the fact that too many end items werestocked in the old situation, which caused problemswith inventory control and shelf lives of products.The number of obsolete products is thus reduced aswell as the inventory costs. Normally, producing onorder could result in less utilisation of capacity.However, here aggregate demand and capacityneeded is quite stable. Moreover, some orders havea longer lead time which enables a smooth produc-tion plan and high utilisation.

Another result of the decisions with respect tothe decoupling point relates to a better knowledgeof the market and the production capabilities andtheir interrelationship. Gathering informationabout products, the demand and the patterns indemand and orders, and lead times, gives a lot ofinformation not yet available to the company inthat way. This opens up discussion regarding thepro"tability of certain articles or the possibilitiesfor using the same recipes for more products. Soa start is made with discussing the marketing andmarket strategy. In such a discussion productioncapabilities play an important role.

5.6. Discussion of the case

The case teaches us that the frame helps us todetect the relevant factors for locating the decoup-ling point and to decide which products should bemade to order or which made to stock. It is clear


that the case has some speci"c elements in it, suchas the possibility to stock intermediate productsand the way capacities in subsequent stages arebalanced.

In the description of the case no attention hasbeen paid to planning and scheduling. Some com-ments can be made, however. Due to the restric-tions made regarding the minimum batch sizes andthe production in larger quantities of product forintermediate storage in silos, the capacity in theprocessing stage will be used e!ectively. The capa-city in the granulation and packaging stages is suchthat the time needed for set-ups leaves enoughspare capacity. Most important, however is thatdemand is stable in terms of the need for capacity.Still, planning and scheduling will be more impor-tant than it was previously and more interactionbetween production and sales department will benecessary. However, that can only be advantageous.

6. Conclusion

This paper develops a frame for a decision, man-agers face in food processing industries: whichproducts have to be made to stock and whichproducts have to be made to order. To supportsuch decisions the general decoupling point con-cept has been adapted to the speci"c characteristicsof the food processing industry. The frame o!ersa systematic means for food processing companiesto "nd the important in#uencing factors in themarket and in their production system. This resultsin a means for making this kind of decision as isillustrated in the case. An important insight fromthis paper is that, at least in the food processingindustry, the decoupling point theory can betransformed into an applicable decision aid formanagers. This paper also contributes to ourknowledge in applying this theory.

The paper shows some problems in decouplingpoint theory such as the development of generalapplicable rules for (changing) the location of thedecoupling point. In fact, balancing the diversefactors in#uencing the location of the DecouplingPoint (both in a qualitative and quantitative sense)is missing in the original writings of Hoekstra andRomme as well. In this particular case, the frame

supported us to develop appropriate decision rules.Other cases are needed to elaborate these rulesfurther for a decision logic for the positioning of thedecoupling point in the food processing industry.We think that progressing along the lines as putforward in this paper will unveil such a logic whichis, at least in the food processing industries, gener-ally applicable.

Further research should also be directed towardsthe consequences of changing the position of thedecoupling point. These consequences apply to theplanning and scheduling (how to combine maketo-stock and make-to-order) and to the capabilitiesof the production systems (see also [4]). Otherareas of attention are the organisational arrange-ments such as the relation between planning,production and marketing and the #ow of informa-tion. Experiences in case studies suggest thatbesides the problem of "nding a balance betweenthe factors in#uencing the position of the decoup-ling point, a major point of concern in implemen-ting chances is to overcome organisational andcultural barriers.

References

[1] M.T.G. Meulenberg, J. Viaene, Changing food marketingsystems in western countries, in: W.M.F. Jongen, M.T.G.Meulenberg (Eds.), Innovation of Food ProductionSystems: Product Quality and Consumer Acceptance,Wageningen Press, Wageningen, The Netherlands, 1998,pp. 8}36.

[2] S. Hoekstra, J. Romme (Eds.), Integral Logistic Structures:Developing Customer-oriented Goods Flow, McGraw-Hill, London, 1992.

[3] J.C. Wortmann, D.R. Muntslag, P.J.M. Timmermans,Customer Driven Manufacturing, Chapman & Hall,London, 1997.

[4] J. Olhager, J. Wikner, A framework for integrated materialand capacity based master scheduling, in: A. Drexl, A.Kimms (Eds.), Beyond Manufacturing Resource Planning(MRP II): Advanced Models and Methods for ProductionPlanning, Springer, Berlin, 1998, pp. 3}20.

[5] S.F. Bolander, Materials management in the process in-dustries, American Production and Inventory Control So-ciety 1980 Conference Proceedings, 1980, pp. 273}275.

[6] S.G. Taylor, S.M. Seward, S.F. Bolander, Why the processindustries are di!erent, Production and InventoryManagement Journal 22 (4) (1981) 9}24.

[7] J.C. Fransoo, W.G.M.M. Rutten, A typology of produc-tion control situations in process industries, International


Journal of Operations & Production Management 14 (12)(1994) 57}72.

[8] P. van Dam, Scheduling packaging lines in the processindustry, Ph.D. Thesis, University of Groningen, Gronin-gen, 1995.

[9] J.P. van Dam, G.J.C. Gaalman, G. Sierksma, Scheduling ofpackaging lines in the process industry: An empiricalinvestigation, International Journal of Production Econ-omics 30}31 (1993) 579}589.

[10] W. van Wezel, D.P. van Donk, Scheduling in food pro-cessing industries: preliminary "ndings of a task orientedapproach, in: J.C. Fransoo, W.G.M.M. Rutten (Eds.), Sec-ond International Conference on Computer IntegratedManufacturing in the Process Industries } Proceedings,BETA, Eindhoven, The Netherlands, 1996, pp. 545}557.


make to stock or make to order: the decoupling point in the food processing industries

Documents