literature review
sindhu16
MassCustomizationliteraturereview.pdf
*Corresponding author. Tel.: 54-2322-48-1064; fax: 54-2322- 48-1050. E-mail address: [email protected] (G. Da Silveira).
Int. J. Production Economics 72 (2001) 1}13
Mass customization: Literature review and research directions
Giovani Da Silveira��*, Denis Borenstein�, FlaH vio S. Fogliatto� �IAE, Universidad Austral, Casilla de Correo, No. 49, Mariano Acosta S/N, Pilar, Buenos Aires, Argentina
�PPGA, Universidade Federal do Rio Grande do Sul, Brazil �PPGEP, Universidade Federal do Rio Grande do Sul, Brazil
Received 12 May 2000; accepted 23 May 2000
Abstract
Mass customization relates to the ability to provide individually designed products and services to every customer through high process #exibility and integration. Mass customization has been identi"ed as a competitive strategy by an increasing number of companies. This paper surveys the literature on mass customization. Enablers to mass customiz- ation and their impact on the development of production systems are discussed in length. Approaches to implementing mass customization are compiled and classi"ed. Future research directions are outlined. � 2001 Elsevier Science B.V. All rights reserved.
Keywords: Mass customization; Customer-driven manufacturing; User involvement; Agile manufacturing
1. Introduction
Mass customization relates to the ability to pro- vide customized products or services through #ex- ible processes in high volumes and at reasonably low costs. The concept has emerged in the late 1980s and may be viewed as a natural follow up to processes that have become increasingly #exible and optimized regarding quality and costs. In addi- tion, mass customization appears as an alternative to di!erentiate companies in a highly competitive and segmented market.
In this paper, we present a literature review on mass customization (MC). Our main objective is to provide a framework to understand the several
developments that emerged in the literature in the past 10 years. We also point to future research directions, based on the current state-of-the-art of the subject. In view of the expanding number of articles and books dealing with MC, there is clear need for a research agenda developed based on existing gaps in the study of MC.
We developed a framework for presenting a sur- vey where MC is deployed from concept to practice in four sections. We start by conceptualizing MC. We want to explore the extent at which theoretical MC concepts describe a production strategy that may be indeed pursued in practice. We then look at the di!erent levels at which MC may be imple- mented. In other words, we classify levels of indi- vidualization that may be provided to customers. After, we move to more applied matters, listing a number of factors that, according to several authors, may lead to a successful implementation of MC. Finally, we discuss in length the enabling
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 7 9 - 7
processes and methodologies to MC implementa- tion.
There are two main contributions here. First, this article presents a comprehensive guide that should help researchers to screen the vast MC literature in search of references on speci"c topics. Through a structured framework, seemingly unconnected aspects of MC are brought together and explored in enough detail to provide a useful introduction to the subject. Second, we set a research agenda cover- ing a variety of important and unexplored facets of MC that should motivate both academics and practitioners to further explore the subject. Despite the increasing attention it has been receiving in the literature, MC is still a novel concept lacking more extensive development. While there is little conten- tion on theoretical aspects such as the MC concept, objectives and justi"cation, the debate over more speci"c and often practical questions remain some- what inconclusive.
2. Mass customization concept
Mass customization (MC) can be de"ned either broadly or narrowly. The broad, visionary concept was "rst coined by Davis [1] and promotes MC as the ability to provide individually designed prod- ucts and services to every customer through high process agility, #exibility and integration [2}4]. MC systems may thus reach customers as in the mass market economy but treat them individually as in the pre-industrial economies [1]. MC systems are positioned below the main diagonal of Hayes and Wheelwright's [5] product}process matrix, i.e. having medium to high-volume process types such as manufacturing cells or assembly lines that are able to deliver the high product varieties usually associated to functional or "xed-type operations.
Many authors propose similar but narrower, more practical concepts. They de"ne MC as a system that uses information technology, #exible processes, and organizational structures to deliver a wide range of products and services that meet speci"c needs of individual customers (often de"ned by a series of options), at a cost near that of mass- produced items [4,6}9]. In any case, MC is seen as a systemic idea involving all aspects of product sale,
development, production, and delivery, full-circle from the customer option up to receiving the "nish- ed product [6,10].
The justi"cation for the development of MC systems is based on three main ideas [4,7,11,12]. First, new #exible manufacturing and information technologies enable production systems to deliver higher variety at lower cost. Second, there is an increasing demand for product variety and custom- ization (according to Kotler [13], even segmented markets are now too broad as they no longer permit developing niche strategies). Finally, the shortening of product life cycles and expanding industrial competition has led to the breakdown of many mass industries, increasing the need for pro- duction strategies focused on individual customers.
3. Levels of mass customization
Determining the level of individualization char- acterizing truly mass-customized products seems to be a major point of contention in the MC debate. Purists may attribute the MC concept only to products that contemplate all requirements made by individual customers. Pragmatists suggest MC to be simply about delivering products following customer options, independent of the number of options actually o!ered. According to Hart [4] the solution for this contention lies in careful determination of the range in which a product or service can be meaningfully customized, and how individuals make options upon this range. To Westbrook and Williamson [14] successful MC systems should be able to mix true individual- ization with high part variety and standardized processes.
Several authors [15,16] propose a continuous framework upon which MC may be developed; namely, MC can occur at various points along the value chain, ranging from the simple `adaptationa of delivered products by customers themselves, up to the total customization of product sale, design, fabrication, assembly, and delivery. Gilmore and Pine [16] identify four customization levels based mostly on empirical observation: collaborative (designers dialogue with customers), adaptive (stan- dard products can be altered by customers during
2 G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13
T ab
le 1
G en
er ic
le ve
ls of
m as
s cu
st o
m iz
at io
n
M C
ge ne
ri c
le ve
ls M
C ap
p ro
ac he
s [1
6] M
C st
ra te
gi es
[1 5]
St ag
es o
f M
C [1
2] T
yp es
o f
cu st
om iz
at io
n [1
8]
8. D
es ig
n C
ol la
bo ra
ti ve
; tr
an sp
ar en
t P
u re
cu st
om iz
at io
n 7.
F ab
ri ca
ti o
n T
ai lo
re d
cu st
om iz
at io
n 6.
A ss
em b
ly C
us to
m iz
ed st
an d
ar d
iz at
io n
M od
u la
r p
ro d
uc ti
o n
A ss
em b
lin g
st an
da rd
co m
p on
en ts
in to
u n
iq u
e co
n "g
u ra
ti o
n s
5. A
dd it
io na
l cu
st o
m w
o rk
P o
in t
o f
d el
iv er
y cu
st o
m iz
at io
n P
er fo
rm in
g ad
d it
io na
l cu
st o
m w
o rk
4. A
dd it
io na
l se
rv ic
es C
us to
m iz
ed se
rv ic
es ;
pr ov
id in
g q
ui ck
re sp
on se
P ro
vi d
in g
ad d
it io
n al
se rv
ic es
3. P
ac ka
ge an
d d
is tr
ib u
ti o
n C
os m
et ic
S eg
m en
te d
st an
d ar
d iz
at io
n C
us to
m iz
in g
p ac
ka gi
n g
2. U
sa ge
A da
pt iv
e E
m be
d de
d cu
st om
iz at
io n
1. St
an da
rd iz
at io
n P
u re
st an
d ar
di za
ti o
n
use), cosmetic (standard products are packaged specially for each customer), and transparent (prod- ucts are adapted to individual needs). Lampel and Mintzberg [15] de"ne a continuum of "ve MC strategies (and therefore levels) involving di!erent con"gurations of process (from standard to cus- tomized), product (from commodities to unique) and customer transaction (from generic to person- alized). A recent study provided empirical evidence of these levels [17]. Pine [12] suggests "ve stages of modular production: customized services (standard products are tailored by people in marketing and delivery before they reach customers), embedded customization (standard products can be altered by customers during use), point-of-delivery customiz- ation (additional custom work can be done at the point of sale), providing quick response (short time delivery of products), and modular production (standard components can be con"gured in a wide variety of products and services). Spira [18] devel- ops a similar framework with four types of custom- ization: customized packaging, customized services, additional custom work, and modular assembly. The combination of these frameworks leads to eight generic levels of MC, ranging from pure customization (individually designed products) to pure standardization; these levels are presented in Table 1.
Design is the top level in Table 1 and refers to collaborative project, manufacturing and delivery of products according to individual customer pref- erences (e.g. residential architecture [15]). Level 7 (fabrication) refers to manufacturing of customer-tailored products following basic, pre- de"ned designs (e.g. Motorola's Bandit pager [3]). Level 6 (assembly) deals with the arranging of modular components into di!erent con"gurations according to customer orders (e.g. Hewlett- Packard products [19]). In levels 5 and 4, MC is achieved by simply adding custom work (e.g. Ikea's furniture [1]) or services (e.g. Burger King's ham- burgers [1]) to standard products, often at the point of delivery. In level 3, MC is provided by distributing or packaging similar products in di!er- ent ways using, for example, di!erent box sizes according to speci"c market segments (e.g. Wal-Mart's peanuts [16]). In level 2, MC occurs only after delivery, through products that can be
G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13 3
adapted to di!erent functions or situations (e.g. Lutron's lighting systems [16]). Finally, level 1 refers to Lampel and Mintzberg's [15] pure standardization, a strategy that can still be useful in many industrial segments.
4. Success factors of mass customization systems
The success of MC systems depends on a series of external and internal factors. The existence of these factors justi"es the use of MC as a competitive strategy and supports the development of MC sys- tems. The six factors most commonly emphasized in the literature are presented next. Factors 1 and 2 are primarily market-related factors. Factors 3}6 are primarily organization-based factors.
1. Customer demand for variety and customization must exist. The need to deal with increasing cus- tomer demand for innovative and customized products is the fundamental justi"cation for MC [2,20,21]. The success of MC depends on the balance between, on the one hand, the potential sacri"ce that customers make for MC products (i.e. how much they will pay and wait for the delivery of mass-customized products [21,22]) and, on the other hand, the company's ability to produce and deliver individualized products within an acceptable time and cost frame.
2. Market conditions must be appropriate. According to Kotha [7], a company's ability to transform MC potential into actual competitive advantage greatly depends on the timing of this develop- ment. In other words, being the "rst to develop an MC system can provide substantial advant- age over competitors, since the company may get well entrenched in this position, and start being seen by people as innovative and cus- tomer-driven.
3. Value chain should be ready. MC is a value chain-based concept. Its success depends on the willingness and readiness of suppliers, distribu- tors, and retailers to attend to the system's demands. The supply network must be at close proximity to the company to deliver raw materials e$ciently [19,21]. Most important, manufacturers, retailers, and other value chain
entities must be part of an e$ciently linked information network [21,23}25].
4. Technology must be available. The implementa- tion of advanced manufacturing technologies (AMTs) is fundamental to enable the develop- ment of MC systems [2,20,21,26,27]. One could argue that the very concept of MC appeared only after some companies were able to success- fully integrate a series of information and pro- cess #exibility technologies. MC is one of the best opportunities o!ered by coordinated imple- mentation of AMTs and information technology (IT) across the value chain.
5. Products should be customizable. Independent units that can be assembled into di!erent forms compose a modular product [19]. Successful MC products must be modularized, versatile, and constantly renewed. Although modularity is not the fundamental characteristic of MC (true MC products are individually made), it enables simpler and lower-cost manufacturing of prod- ucts with similar e!ectiveness if compared to true customization. Also, MC processes need rapid product development and innovation capabilities due to typical short life cycles pre- sented by MC products [2,20].
6. Knowledge must be shared. MC is a dynamic strategy and depends on the ability to translate new customer demands into new products and services. To achieve that, companies must pursue a culture that emphasizes knowledge creation and distribution across the value chain. That requires the development of dy- namic networks [2] along with manufacturing and engineering expertise [28], and in-house development of new product and process tech- nologies [7].
These factors have direct practical implications. First, they corroborate the idea that MC is not every company's best strategy, for it must conform to speci"c market and customer types. Second, they assert the complexity involved in MC implementa- tion. In other words, MC implementation involves major aspects of operations including product con"guration, value chain network, process and information technology, and the development of a knowledge-based organizational structure.
4 G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13
Table 2 MC enablers and related success factors
Enablers Related success factors (organization-based)
Processes and methodologies Agile manufacturing Knowledge Supply chain management Value chain Customer-driven design and manufacturing Customizable products Lean manufacturing Value chain
Enabling technologies Advanced manufacturing technologies Technology, customizable products Communication and networks Technology, knowledge
5. Enablers of mass customization implementation
MC enablers are the methodologies and tech- nologies that support the development of the organization-based factors (i.e. factors from 3 to 6) described above (Table 2). This section is divided in three subsections discussing (i) processes and methodologies enabling MC, (ii) technologies enabling MC and (iii) how technologies support information transfer, that is perhaps the major implementation problem with MC.
5.1. MC processes and methodologies
MC processes and methodologies address the organizational and cultural aspects of implemen- ting an MC system. They concern the main elements of a manufacturing strategy supporting the development of successful MC systems, capable of providing the elements cited in the previous section. Analysis of the literature points to the existence of at least four main business practices relating to the MC concept: agile manufacturing [29,30], supply chain management [3,31], cus- tomer-driven design and manufacture [1,7,18], and lean manufacturing [32,33]. Agile manufacturing has been de"ned as the abil-
ity to thrive and prosper in a competitive environ- ment of continuous and unanticipated change to respond quickly to rapidly changing markets driven by customer-based valuing of products [34]. Agile manufacturing is characterized by the con- scious usage of a changing environment as a means to be pro"table. While a #exible manufacturer is
characterized by a reactive adaptation behavior (waiting for a change to occur to act), an agile manufacturer has a proactive behavior [35]. DeVor et al. [36] identify the main strategic dimensions of agile manufacturing as (i) value- based strategies that enrich customers, focusing on delivering value; (ii) cooperating to enhance com- petitiveness; (iii) organizing to master change and uncertainty; (iv) leveraging the impact of people and information.
These dimensions lead to the concept of internal and external agility [30]. Internal agility may be viewed as the ability to quickly respond to market and customer demands for new products and prod- uct features. That requires re-programmable, re-con"gurable and continuously changeable pro- duction systems able to operate economically with very small lot sizes [37]. Researchers have also discussed the cultural aspects associated with agil- ity. Owen and Kruse [30], for example, point out that a true learning organization is necessary for agility to succeed.
External agility is associated with the idea of virtual enterprises [38}40]. A virtual enterprise consists of several individual companies linked in a collaborative e!ort to design high-quality and customized products [41]. Virtual organizations have the following main characteristics [40]: product orientation, team-collaboration style, short-term relationships between individuals, speed, and #exibility. This organizational model is enabled by the availability of sophisticated in- formation technologies and telecommunication systems [42].
G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13 5
Supply chain management concerns the co-ordi- nation of resources and the optimization of activ- ities across the value chain to obtain competitive advantages [43]. As previously discussed, e$cient supply chain management is one of the key success factors in systems. Eastwood [3], Feitzinger and Lee [19], Lau [20], Kotha [7], and Moad [44] describe how improving supply management pro- vides organizational coordination required in MC systems. Such conditions include: (i) development of an interconnected information network involv- ing a selected group of trained suppliers; (ii) suc- cessful balance of low stocks with high delivery service; (iii) design of innovative products with active collaboration of suppliers; and (iv) cost- e!ective delivery of the right product to the right customer at the right time. Customer-driven design and manufacture is in the
core of MC systems. Jagdev and Browne [37] de"ne this business practice as `to actively consider the market trends in general and individual cus- tomer requirements in particular during the design, manufacturing and delivery of the productsa. Some authors call this practice `One-of-a-Kind Produc- tiona (OKP [45]). The application of customer- driven practices in MC systems aims fundamentally at (i) providing conditions for the customer to initi- ate the design process of a product, and (ii) building an infrastructure to develop new products driven by the market. The number of di!erent product variants (Bally Engineered Structures, Inc., for example, have about 100,000 items in catalogue [46]) is a consequence of implementing customer- driven ideas. Section 6.2 describes a sequence of steps to establish a consumer}manufacturer com- munication link within the consumer-driven design and manufacture principles. Lean manufacturing is an e$cient way to satisfy
customer needs while giving producers a competi- tive edge [47]. The MC production addresses four elements of lean production: product development, the chain of supply, shop #oor management, and after-sales services [48]. For a successful implemen- tation of an MC production system, it is essential (i) to de"ne value based on the customer, (ii) to concentrate in the activities that create value and to eliminate all wastes, in all production steps, and, "nally, (iii) to reorganize the value-creating
activities into e$cient processes, without interrup- tions and incorporating production variant at high levels.
5.2. MC enabling technologies
Main enabling technologies supporting MC are AMTs, such as computer numeric control (CNC) and #exible manufacturing systems (FMS), and communication and network technologies such as computer-aided design (CAD), computer-aided manufacturing (CAM), computer integrated manu- facturing (CIM), and electronic data interchange (EDI) [26,27,49]. As previously mentioned, many researchers [2,7] consider such technologies funda- mental to MC implementation.
The use of AMTs is justi"ed by their inherent capability to alter the economies of manufacturing and remove barriers to product variety and #exibil- ity [50]. These technologies enable the develop- ment of factories that can exploit the bene"ts of such fundamental MC attributes as agility and #exibility.
Case examples such as the NBIC [7], Motorola [3], and Perkins [51] stress the important role of AMT in MC system development. NBIC employed CAD/CAM, advanced computer-controlled ma- chines, and robots in implementing their MC manufacturing system. Motorola used CIM-related technologies (such as Cartesian and gantry robots) to implement two MC factories. Perkins based their MC system on a hybrid CAD/CAE (com- puter-aided engineering) system with #exible manufacturing assembly lines.
The main motivation behind the extensive use of communications and networks based on informa- tion technology is to provide direct links between work-groups (e.g. design, analysis, manufacturing, and testing), and to improve the response time to customer requirements. The communication and network technologies are tools to integrate pre- viously isolated components of a productive chain into coherent and coordinated competitive weapons.
Bally [46] and Betchel [50] are examples of the use of intensive information technology to implement MC concepts. Bally employed ad- vanced information technology, based on arti"cial
6 G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13
intelligence methods, to move from CIM to CDIN, a computer-driven intelligence network; this system links their sales representatives and suppliers in one single information network. Betchel developed an advanced information system consisting of an in- tegrated collection of engineering, procurement, construction, and project management software modules.
5.3. Enabling technologies at work: Information transfer
The e$ciency in information transfer from cus- tomers to manufacturers determines largely the success of an MC program [42]. In MC programs, customer demands regarding a product are cap- tured and transmitted to a production unit, where a product tailored to meet those demands is manu- factured. Agents of information transfer are the manufacturer and its customers. The manufacturer de"nes to what extent customers may customize their order; customers feed back the information on their choice of design elements. The customer} manufacturer interface is uniquely de"ned accord- ing to the company developing and implementing an MC program. The following sequence of steps attempts at describing activities involved in estab- lishing a customer}manufacturer communication link: (i) de"ning a catalogue of options to be o!ered to customers; (ii) collecting and storing information on customer choices; (iii) transferring data from retail to manufacturer; and (iv) translating cus- tomer choices into product design features and manufacturing instructions.
The degree at which products are customized (see the discussion on MC levels above) is unlikely to exclude any of the steps above; it de"nes, how- ever, the volume of information transferred in each step. In the following paragraphs, a literature sur- vey on customer involvement in MC is presented using the four steps above as guidelines. Step 1 } Dexning a catalogue of options to be
owered to customers. The catalogue of options of- fered to customers de"nes the degree of customiz- ation of a product. Highly customized products present extensive catalogue of options, covering most of their relevant features. Medium- and low- customized products o!er choices that are more
restricted to customers. Some products are o!ered in models developed based on the analysis of cus- tomer's past demands. In other words, customer may choose from several pre-determined models with design features likely to match their needs. This corresponds to a very low degree of customiz- ation, in which customer interaction in the product design is indirect. Table 3 lists examples for these customization options. It is important to note that the o!er of choices, although essentially customer driven, must be coherent with the manufacturer's technological development [52]. Step 2 } Collecting and storing information on
customer choices. There is no proli"c literature on the collection and storage of information from cus- tomers. As expected, approaches for data collection are developed to attend speci"c MC situations. Data on customer choices may be gathered by a store employee or sales representative trained to guide him or her through the decision process [1,18,21,46,53,54], or may be collected using a com- puter interface with minimum human interference [16]. In other situations, customer and designer jointly develop a project from scratch [1,55]. In any case, it is implicit that customers must be o!ered an easy-to-handle set of options to select. Information is commonly stored on order sheets [21] or elec- tronically, using a computer system [1,16,18,46,53]. Genetic algorithms and autonomous agents are also presented as facilitators in the data acquisition process [56]. Step 3 } Transferring data from store to manufac-
turer. In all reported cases, orders are sent from store to manufacturer by FAX [21,46] or computer link [1,16,18,53]. More recent cases present the Internet as a means to link store to manufacturing, e.g. automobile (Fiat, Pontiac), and textile indus- tries (Levi's). Information on customer preferences are then entered in a computer system that gener- ates a product ID, such as a bar code, used to track the product throughout the manufacturing stages. Step 4 } Translating customer choices into product
design features and manufacturing instructions. In most reported cases, speci"cations on design elements are fed into CAD and CAM systems and then converted into production instructions [7,18,57]. It is evident that the success of MC im- plementations is heavily dependent on the existence
G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13 7
Table 3 MC examples and their main information-handling characteristics
Product Translating customer choice
Data transfer Data storing Data gathering Catalogue of options
Cable control system * * * * High Diesel engines * * * * High Printers Forecasting * * * Low Bicycles CAD/CAM FAX Order sheet Store employee High Refrigeration system CAD FAX Comp. interface Sales rep. Low Computers Forecasting Computer Comp. interface Sales rep. Low Lighting controls CAD/CAM Computer Comp. interface Sales rep. High Insurance * * * Sales rep. Low Power generator CAD/CAM * * * Low Houses CAD/CAM Computer Comp. interface Sales rep. Low Eyeglasses * * Comp. interface Store employee/ computer High Shoes CAD/CAM Computer Comp. interface Store employee/ computer Medium Pagers CAD/CAM * * * High
of a computerized manufacturing environment. Therefore, CAD and CAM systems are key when attempting any MC strategy. This is expected since, in essence, MC relies on #exibility and quick responsiveness. CAD systems allow customer- driven design changes to be implemented and de- ployed into production instructions in due time; CAM systems handle the diversity of parts ordering while maximizing machine use.
Table 3 lists examples of mass-customized prod- uct along with their main information-handling characteristics.
6. Research agenda
Future research on MC should focus on the formulation of methodologies that enable rapid recon"guration of existing organizational struc- tures and processes into a mass-customized pro- duction system. In this sense, further developments in MC research tend to point towards more applied issues.
6.1. Customization level assessment
As seen earlier, the literature provides a series of frameworks describing di!erent levels of customiz- ation that may be adopted in practice [12,15,18].
However, these studies do not provide enough knowledge on how to determine the appropriate level of customization for a speci"c product or service. High-level customization involves signi"- cant competitive bene"ts but also operational costs. Determining the right level of customization depends on appropriate analysis of customer requirements and existing operational capabilities.
An important contribution to the MC literature could come in the form of a methodology for deter- mining the appropriate level of MC for a product or service, e.g. customized design, fabrication, or assembly. Such methodology would most likely address three important problems: (a) measur- ing the value customers provide to a level of customization, (b) measuring the system's ability to deliver a level of customization, and (c) compar- ing and combining these seemingly con#icting measures.
One idea is to forge that methodology from earlier methods such as quality function deploy- ment (QFD [58]) and the importance}performance matrix (IPM [59]). QFD in conjunction with suit- able sample survey techniques could be used to identify and rank customizable features in products and services that could potentially meet customers' demand. IPM in conjunction with selected #exibility indices [60] may be used in measuring the ability to deliver the required level of customization.
8 G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13
Development of such methodology requires (i) exploratory, case-based research to identify the main issues involved with assessing customer requirements and process capabilities relating to customization and (ii) theoretical or action research to propose and validate such a methodology
6.2. Mass customization in services
The lack of studies dealing with MC in service operations is perhaps one of the main gaps in the current MC literature. The existing research is still largely focused on manufacturing operations, specially batch industries. Table 3 provided a sum- mary of examples of mass-customized products in the literature; in 13 examples, only one came from the service industry (insurance). In addition, all examples in the discussion on MC levels referred to manufacturing goods, with the possible exception of Burger King.
The need to develop research on mass-cus- tomized services becomes even greater if we con- sider the many di!erences between manufacturing and services operations, and the implications these di!erences may have in the design of MC systems. In comparison to manufacturing, service opera- tions are (i) more labor-intensive, (ii) have greater customer involvement, (iii) are more sensitive to quality errors, (iv) have tighter delivery times, (v) are unable to rely on inventories to adjust to demand #uctuations, and (vi) are more dependent on information reliability [61,62]. Besides, service products are intangible, usually transient, and have more subjective qualities than manufac- tured products. All these elements can pose challenges to MC implementation, e.g. how to develop a #exible and skilled workforce, how to deliver products that match closely the customer requirements, how to guarantee quality and safety despite changes in service parameters, and how to promise short delivery times for mass-customized services.
Research on MC in services could be either the- oretical (e.g. matching categories of manufacturing versus service industries, and how these relate to MC elements), exploratory (e.g. survey of service industries to identity practices), or descriptive (e.g. case studies of service companies).
6.3. Information management
An MC system is highly dependent on well- designed information systems that provide direct links between internal work-groups, such as manu- facturing, design, and testing, and external work- groups, represented by suppliers and customers. However, there is a void in the literature on how to implement the information management processes required in MC.
It is possible to identify the following opportuni- ties for research topics involving this important issue:
� Design of an e!ectively decentralized, or multi- agents, control architecture for MC systems. This control system is composed of autonomous components. Its implementation will reduce the complexity, increase #exibility, and enhance fault tolerance needed to successfully implement a team of independent but cooperating pro- ducers.
� Application of modeling architectures in MC environment. In particular, enterprise modeling, such as open systems architectures for computer integrated manufacturing (CIMOSA) and archi- tecture for integrated information systems (ARIS) [63]. Enterprise modeling encompasses modeling, analysis, design, and implementation of integrated information systems [64]. The con- struction of an enterprise model for an MC "rm should embrace overall system architecture, product design, project management, software speci"cation, and establishing the data model for data base design [65].
� Development of an information management infrastructure for MC systems based on the integration of di!erent standards or tools: Inter- net, STEP, and object-oriented paradigm. This infrastructure may enable the software tools for MC systems.
6.4. Quality control and monitoring
On-line and o!-line quality control practices are applied for measuring the performance of processes or products. That is usually accomplished by monitoring the behavior of selected quality
G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13 9
characteristics (QCs) over time. For that purpose, many strategies have been presented in the litera- ture, most notably those proposing the use of stat- istical control charts [66] and poka-yoke devices [67]. Two key points determine the success of most quality control schemes: (i) de"nition of rel- evant QCs to be monitored; these characteristics must be chosen to re#ect customers' quality demands, and (ii) availability of data on the QCs to be monitored from which a reference behavior may be inferred.
MC systems are characterized by single product lots, which are unlikely to be repeated over time. Consequently, traditional quality control schemes such as statistical control charts, which operate based on periodical checking on selected QCs, would not be easily adapted to MC environments. The problem rises in complexity as one notes that a new set of QCs is de"ned whenever a product is customized. It is natural to expect most relevant QCs to remain unchanged as customization takes place, since they relate to basic operational functions of products. However, new explicit QCs are likely to arise, which implies in their identi"ca- tion and establishing of a monitoring scheme for them.
It seems clear from the discussion above that quality control issues should be taken into account when deciding upon product customization; they are likely to bind the level of customization admis- sible to products and services, if these are to comply to current quality standards. It is important to note that the current MC literature lacks any in depth study on how to assure quality in mass-customized products.
One possible approach to ease the burden of product quality monitoring, and therefore allow for customized products, is to guarantee on-target product QCs mostly through monitoring of pro- cesses. For that purpose, one shall determine the mathematical relationships between product-re- lated QCs and those that are process-related. One idea is to ground product development on methods such as QFD and favor intensive use of statistical tools such as design of experiments [68], and ro- bust parameter optimization [69]. QFD allows for qualitative determination of relationships between product QCs and process QCs. Such relationships
could then be mathematically measured using de- sign of experiments. Finally, the number of process QCs to be monitored could be reduced through robust parameter optimization (based on the prop- osition that robust processes require less intensive monitoring). Development of such methodology requires (i) theoretical research to propose method- ological steps to guide MC quality optimization studies; (ii) case-based research to test and re"ne the steps proposed in (i).
6.5. Reliability analysis of mass-customized products
The reliability of a component or system is de- "ned as the probability that it will adequately per- form its speci"ed purpose for a speci"ed period of time under speci"ed environmental conditions [70]. To assess such probability, the lifetime of components or systems must be determined from empirical data. Data may be gathered from "eld observation of failures or from lab tests carried under normal environmental conditions or under accelerated stress [71]. In any case, reliability stud- ies tend to be time consuming and, in most practi- cal situations, unavoidable if products or systems are to comply with international reliability stan- dards.
MC products have short life cycles and ever changing characteristics. Reliability analyses of such products would imply an accelerated testing which are, in most instances, very costly. Depend- ing on the product level of customization, reliability testing of one basic product would su$ce for all variants that may arise from its customization; that would be the case when cosmetic or adaptive customization are the manufacturer's choice. However, upon practicing of collaborative or trans- parent customization, grounding of reliability in- ferences on data from basic designs, subjected to several changes after customer intervention, tends to o!er undependable results.
There are many cases in the MC literature deal- ing with products typically subjected to reliability testing. For example, printers in Feitzinger and Lee [19], cable control systems in Owen and Kruse [30], pagers in Eastwood [3], and power gener- ators in Choi and Jarboe [57]. These references do not provide any indication on (i) how reliability
10 G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13
testing of products were performed, or (ii) how reliability constraints in#uenced the choice of cus- tomization level. Clearly, such issues are central in the design of customized products, deserving further research.
Most manufactured MC products are de"ned upon combining pre-determined choices from a "nite set of options. In that context, a relevant contribution to the MC literature could come in the form of a method for assessing product reliabil- ity resulting from di!erent combinations of choices. Such method would probably rely on computer simulation of product operation, since accelerated testing of all possible combinations of choices would be economically infeasible. One idea is to measure the reliability of parts or components de- "ning the product, rather than measuring the prod- uct reliability itself. Whenever a combination of choices generated a new customized product, reliability could be assessed upon analysis of the resulting reliability block diagram [72]. Such approach would require knowledge of the depend- ence among parts in the product under study [72]; combining information from the reliability block design and the dependence evaluation, mathemat- ical models on which simulations could be based would be at hand. Development of such methodo- logy tends to be heavily based on theoretical research, in particular topics dealing with time (and failure) dependent reliability, validated by empiri- cal testing.
7. Conclusions
MC has become an important manufacturing strategy. Agility and quick responsiveness to cha- nges have become mandatory to most companies in view of current levels of market globalization, rapid technological innovations, and intense competi- tion. MC broadly encompasses the ability to pro- vide individually designed products and services to customers in the mass-market economy.
However, MC should not be viewed as a mono- lithic solution. Manufacturing processes are too complex and context sensitive for a single black- box idea to generate #exible, agile, and focused systems. To implement MC it is necessary to
integrate di!erent manufacturing technologies into a structured framework capable of combining human and technological factors.
This paper presents a literature review on MC. The objective is to identify required conditions and situations where MC implementation is suitable. In addition, fundamental principles and concepts in MC theory are thoroughly discussed.
The study reveals that, while there is little contention on theoretical aspects concerning MC concepts and objectives, there are several pending issues regarding its practical implementation. Liter- ature on MC implementation is still incipient. Most claims are drawn from limited case examples or based on educated guesses from authors rather than from hard evidence obtained through exhaustive research. The paper closes presenting some directions for further research.
References
[1] S. Davis, From future perfect: Mass customizing, Planning Review 17 (2) (1989) 16}21.
[2] J. Pine, B. Victor, A. Boyton, Making mass customization work, Harvard Business Review 71 (5) (1993) 108}111.
[3] M. Eastwood, Implementing mass customization, Com- puters in Industry 30 (3) (1996) 171}174.
[4] C. Hart, Mass customization: Conceptual underpinnings, opportunities and limits, International Journal of Service Industry Management 6 (2) (1995) 36}45.
[5] R. Hayes, S. Wheelwright, Linking manufacturing process and product life cycle, Harvard Business Review 57 (1979) 133}140.
[6] M. Kay, Making mass customization happen: Lessons for implementation, Planning Review 21 (4) (1993) 14}18.
[7] S. Kotha, Mass customization: Implementing the emerg- ing paradigm for competitive advantage, Strategic Man- agement Journal 16 (1995) 21}42.
[8] A. Ross, Selling uniqueness, Manufacturing Engineer 75 (6) (1996) 260}263.
[9] A. Joneja, N.-S. Lee, Automated con"guration of paramet- ric feeding tools for mass customization, Computers and Industrial Engineering 35 (3}4) (1998) 463}469.
[10] J. Jiao, M. Tseng, V. Du!y, F. Lin, Product family modeling for mass customization, Computers and Indus- trial Engineering 35 (3}4) (1998) 495}498.
[11] P. Ahlstrom, R. Westbrook, Implications of mass custom- ization for operations management: An exploratory sur- vey, International Journal of Operations and Production Management 19 (3) (1999) 262}274.
[12] J. Pine, Mass customizing products and services, Planning Review 21 (4) (1993) 6}13.
G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13 11
[13] P. Kotler, From mass marketing to mass customization, Planning Review 17 (5) (1989) 10}13.
[14] R. Westbrook, P. Williamson, Mass customization: Japan's new frontier, European Management Journal 11 (1) (1993) 38}45.
[15] J. Lampel, H. Mintzberg, Customizing customization, Sloan Management Review 38 (1996) 21}30.
[16] J. Gilmore, J. Pine, The four faces of mass customization, Harvard Business Review 75 (1) (1997) 91}101.
[17] G. Amaro, L. Hendry, B. Kingsman, Competitive advant- age, customization and a new taxonomy for non make-to- order companies, International Journal of Operations and Production Management 19 (4) (1999) 349}371.
[18] J. Spira, Mass customization through training at Lutron Electronics, Computers in Industry 30 (3) (1996) 171}174.
[19] E. Feitzinger, H. Lee, Mass customization at Hewlett- Packard: The power of postponement, Harvard Business Review 75 (1) (1997) 116}121.
[20] R. Lau, Mass customization: The next industrial revol- ution, Industrial Management 37 (5) (1995) 18}19.
[21] S. Kotha, From mass production to mass customization: The case of the National Industry Bicycle Company of Japan, European Management Journal 14 (5) (1996) 442}450.
[22] C. Hart, Made to order, Marketing Management 5 (2) (1996) 10}23.
[23] M. Haglind, J. Helander, Development of value networks } an empirical study of networking in Swedish manufac- turing industries, Proceedings of the International Confer- ence on Engineering and Technology Management, 1999, pp. 350}358.
[24] J. Kim, Hierarchical structure of intranet functions and their relative importance: using the Analytic Hierarchy Process, Decision Support Systems 23 (1) (1998) 59}74.
[25] J. Magretta, The power of virtual integration: An interview with Dell Computer's Michael Dell, Harvard Business Review 76 (2) (1998) 72}84.
[26] B. Hirsch, K.-D. Thoben, J. Hoheisel, Requirements upon human competencies in globally distributed manufactur- ing, Computers in Industry 36 (1}2) (1998) 49}54.
[27] P. Kanchanasevee, G. Biswas, K. Kawamura, S. Tamura, Contract-net based scheduling for holonic manufacturing systems, Proceedings of the SPIE } The International Society for Optical Engineering 3203 (1999) 108}115.
[28] S. Kotha, Mass-customization: a strategy for knowledge creation and organizational learning, International Jour- nal of Technology Management 11 (7/8) (1996) 846}858.
[29] E. Adamides, Responsibility-based manufacturing, Inter- national Journal of Advanced Manufacturing Technology 11 (1996) 439}448.
[30] D. Owen, G. Kruse, Follow the customer, Manufacturing Engineering 118 (4) (1997) 65}68.
[31] T. Gooley, Mass customization: How logistics makes it happen, Computers and Industrial Engineering 37 (4) (1998) 49}54.
[32] J. Womack, D. Jones, D. Ross, The Machine that Changed the World, Rawson, New York, 1990.
[33] J. Womack, D. Jones, Lean Thinking, Simon & Schuster, New York, 1996.
[34] Iacocca Institute, 21st Century Enterprise Strategy, Vol. 1, Lehigh University Press, Bethlehem, PA, 1991.
[35] R. Gutman, R. Graves, The agile manufacturing enterprise } both a new paradigm and a logical extension of #exible and lean, EAMRI Report ER95-10, Rensselaer Polytech- nic Institute, Troy, NY, 1995.
[36] R. DeVor, R. Graves, J. Mills, Agile manufacturing re- search: Accomplishments and opportunities, IIE Transac- tions 29 (10) (1997) 813}823.
[37] H. Jagdev, J. Browne, The extended enterprise } a context for manufacturing, Production Planning and Control 9 (3) (1998) 216}229.
[38] S. Gutman, R. Graves, K. Preiss, Agile Competitors and Virtual Organizations, Van Nostrand Reinhold, New York, 1995.
[39] I. Shekhar, R. Nagi, Automated retrieval and ranking of similar parts in agile manufacturing, IIE Transactions 29 (10) (1997) 859}876.
[40] L. Song, R. Nagi, Design and implementation of a virtual information system for agile manufacturing, IIE Transac- tions 29 (1) (1997) 839}857.
[41] S. Powell, F. Gallegos, Securing virtual corporations, Information Structure: The Executive's Journal 14 (4) (1998) 34}38.
[42] K. Turowski, A virtual electronic call center solution for mass customization, Proceedings of the 32nd Annual Hawaii International Conference on Systems Sciences, 1999, pp. 152}164.
[43] A. Boyton, B. Victor, J. Pine, New competitive strategies: Challenges to organizations and information technology, IBM Systems Journal 32 (1) (1993) 40}64.
[44] J. Moad, Let customers have it their way, Datamation 41 (April) (1995) 34}39.
[45] J. Wortman, Factory of the future: Towards an integrated theory of one of a kind production, in: B. Hirsch, K. Thoben (Eds.), One of a Kind Production: New Approaches, North-Holland, Amsterdam, 1992, pp. 31}74.
[46] B. Pine, T. Pietrocini, Standard modules allow mass cus- tomization at Bally Engineered Structures, Planning Review 22 (4) (1993), 20}22.
[47] R. Storch, S. Lim, Improving #ow to achieve lean manu- facturing in shipbuilding, Production Planning and Con- trol 10 (2) (1999) 127}137.
[48] H. Warnecke, M. HuK ser, Lean production, International Journal of Production Economics 41 (1995) 37}43.
[49] W. King, IT-enhanced productivity and pro"tability, Information Systems Management 15 (3) (1998) 70}72.
[50] J. Meredith, Automating the factory, International Journal of Production Research 25 (10) (1987) 1493}1510.
[51] G. Vasilash, Mass customization at Perkins: An engine with one-trillion possibilities, Automotive Manufacturing and Production 109 (2) (1997) 42}44.
[52] E. McCarthy, Mass-customizing client service through high-tech, high-touch communications, Journal of Finan- cial Planning 10 (3) (1997) 58}63.
12 G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13
[53] R. Beaty, Mass customization, Manufacturing Engineer- ing 75 (5) (1996) 217}220.
[54] H. Martin, Mass customization at personal lines insurance center, Planning Review 21 (4) (1993) 27}56.
[55] J. Kubiak, A joint venture in mass customization, Plann- ing Review 22 (4) (1993) 19}25.
[56] B. Fulkerson, A response to dynamic change in the market place, Decision Support Systems 21 (3) (1997) 199}214.
[57] K. Choi, T. Jarboe, Mass customization in power plant design and construction, Power Engineering 100 (1) (1996) 33}36.
[58] L. Cohen, Quality Function Deployment } How to Make QFD Work for You, Addison-Wesley, Reading, MA, 1995.
[59] N. Slack, The importance-performance matrix as a deter- minant of improvement priority, International Journal of Operations and Production Management 14 (5) (1994) 59}75.
[60] A. Sethi, S. Sethi, Flexibility in manufacturing: A survey, International Journal of Flexible Manufacturing Systems 2 (1990) 289}328.
[61] R. Murdick, B. Render, R. Russel, Service Operations Management, Allyn and Bacon, Needham Heights, MA, 1990.
[62] C. Voss, Operations Management in Service Industries and the Public Sector: Test and Cases, Wiley, Chichester, 1985.
[63] G. Doumeingts, D. Chen, State-of-the-art on models, architectures and methods for CIM systems design, in:
G. Olling, F. Kimura (Eds.), Proceedings of the IFIP TC5/WG 5.S, Eighth International PROLAMAT Conference: Human Aspects in Computer Integrated Manufacturing, Elsevier Science, Amsterdam, 1992, pp. 27}40.
[64] F. Venadat, Enterprise Modeling and Integration: Prin- ciples and Applications, Chapman & Hall, New York, 1997.
[65] M. Aguiar, R. Weston, A model driven approach to enter- prise modeling, International Journal of Computer Integ- rated Manufacturing 8 (1995) 210}224.
[66] D.C. Montgomery, Introduction to Statistical Quality Control, 3rd Edition, Wiley, New York, 1996.
[67] S. Shingo, Zero Quality Control: Source Inspection and the Poka-Yoke System, Productivity Press, Cambridge, 1986.
[68] D.C. Montgomery, Design and Analysis of Experiments, 3rd Edition, Wiley, New York, 1991.
[69] S. Park, Robust Design and Analysis for Quality Engineer- ing, Chapman & Hall, London, 1996.
[70] L. Leemis, Reliability: Probabilistic Models and Statistical Methods, Prentice-Hall, Englewood Cli!s, NJ, 1995.
[71] W. Nelson, Accelerated Testing: Statistical Models, Test Plans and Data Analyses, Wiley, New York, 1990.
[72] E.A. Elsayed, Reliability Engineering, Addison-Wesley, Reading, MA, 1996.
G. Da Silveira et al. / Int. J. Production Economics 72 (2001) 1}13 13
