You are on page 1of 4

CIRP Annals - Manufacturing Technology 59 (2010) 175–178

Contents lists available at ScienceDirect

CIRP Annals - Manufacturing Technology


jou rnal homep age : ht t p: // ees .e lse vi er. com/ci rp/ def a ult . asp

Design for mass personalization


M.M. Tseng (1)a,*, R.J. Jiao b, C. Wang a
a
Advanced Manufacturing Institute, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong
b
The G.W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA 30332-0405, USA

A R T I C L E I N F O A B S T R A C T

Keywords: With pervasive connectivity of the Internet, personalization has become increasingly accepted for digital
Design
products. As opposed to customization, which emphasizes on meetings explicit requirements of defined
Product development
market segments, personalization aims at effectively and efficiently satisfying individual needs. By
Personalization
considering customers as individuals, implicit characteristics such as personal taste, traits, innate needs
and experience become important integral parts of product design. This new dimension for design opens
up many new research issues. Design for mass personalization (DFMP) aims at effectively and efficiently
satisfying customers as individuals by offering personally unique products with positive user experience.
To this end, design is approached through the formulation of a product ecosystem based on a design
platform and active customer participation.
ß 2010 CIRP.

1. Introduction customers themselves either. Then current product life cycle from
product conception, delivery, usage, service, thru end of life
With the growing attention to increase the value added in disposal has not been able to consider customers as individuals. It
manufacturing, mass customization (MC) has attracted much is apparent that not only designers but also customers play crucial
attention as a major strategic initiative from academia and roles in expanding the scope of design from customization to
industry [1]. The prevailing practice of customization is to identify personalization. Towards this end, design for mass personalization
patterns of customer needs with families of products, along with (DFMP) aims at effectively and efficiently satisfying customers as
common building blocks of components, subassemblies, and individuals by offering personally unique products with positive
modules with product fulfilment processes. Thus, various elements user experience.
of design and manufacturing can be reused to match customer
needs with product variety [2,3]. MC has managed to serve 2. Personalization and design
businesses with market niches, but still not to reach the extent of
market-of-one which has been envisioned as a potential driving Personalization is to achieve satisfying each customer as an
force for the next transformation of global economy [4]. individual. Thus, product differentiation is at individual customer
In information systems, personalized user interface has been level, as opposed to customization which differentiates products
popular by using implicit data captured in user interactions. In for market segments. The idea of deriving profitability from, and
essence, personalization is achieved through eliciting customer gaining competitive edges through, differentiating products is not
inherent needs based on one-to-one interactions with the new. It often appears as market segmentation, customer-centric, or
customer. The rapid emergence of Web technologies with expert augmented cognition to exemplify personalization. In terms of
systems techniques has enabled market-of-one strategy in online design, personalization discerns from customization mainly in two
sale and e-service [5]. For example, Amazon.com can recommend dimensions, expanding product design space, and embracing
customers to source books or products that they have not heard of intangible customer experience.
based on their profiles and purchase histories. It provides
customers with unique purchase experience and personalized 2.1. Product design space
product offerings beyond expectation. Will this be possible in new
design technology? Can product development team anticipate and In customization, customers are classified into different market
adapt to customers’ latent needs? Can customers be delighted as segments based on different known customer needs identified by
individuals as suggested by the Kano model [6]? Can personaliza- marking analysis. The customers grouped in the same market
tion be carried out with efficiency? segment receive similar or parameter-based customized products
To begin with, the conventional survey or interview often fails from a pre-defined product family. The configuration mechanism,
to reveal latent customer needs. Furthermore, customers can be product architectures, basic modules and critical parameters have
reluctant to reveal their inner needs, which may not be known by been kept stable within the pre-defined realm of configuration. As
a matter of fact, the essence of design for mass customization
(DFMC) is to configure various product variants through mod-
* Corresponding author. ularity with the commonality embedded in product platforms so as

0007-8506/$ – see front matter ß 2010 CIRP.


doi:10.1016/j.cirp.2010.03.097
176 M.M. Tseng et al. / CIRP Annals - Manufacturing Technology 59 (2010) 175–178

to reuse proven design among product families [2]. On the other 3. Design for mass personalization
hand, in DFMP, customers work with providers to create objects
that may be totally new to fulfil customer needs as long as budget In order to include user experience in unknown design space,
and schedule permit. design for mass personalization (DFMP) needs to synthesize the
Personalization goes beyond configuration-to-order product experience into the innovative process of design and carry it out
lines and needs to explore market potential by leveraging product with efficiency. DFMP includes product ecosystem as the kernel,
functionality with the realization of affective and cognitive needs, design framework as the design context and customer co-creation
hence enabling product differentiation beyond the original set of to drive the process. They are described in the following sections.
product offerings. While customization assumes fixed product
architectures and pre-defined configuration models, personaliza- 3.1. Product ecosystem
tion implies possible changes of the basic design and product
features. The adaptability and changeability of product designs The inter-twined relationships among objects in modern day
becomes essential for personalization. life are inherent and complementary. The products are no longer
Both customization and personalization involve changes of islands of their own to fulfil self-contained isolated functionality.
product design to meet the customer needs. In customization, Everything from digital cameras, media players, car accessories,
individual customers make their selection base on ‘the menu’ mobile phones, and even clothing and running shoes, have an
which consists mainly of a collection of known (and proven) important interdependence with broader user related character-
components, modules and subsystems with given logic to istics (comparing with functional performance related compo-
configure finished products. Computer configuration is an nents of the product) that are related to customer participations. As
excellent example here. Computer manufacturers predict in customers become more connected, products and services are
advance the potential ranges of products that customers would increasingly knitted to a larger ecosystem of interacting objects.
like to purchase. Producers then determine the common denomi- The product is not an isolated incidence, it should be considered as
nators to find the best set of building blocks for customers to mix a part of ecosystems. Looking at the whole sum of interacting
and match to get to their needs. In personalization, the final objects and understanding the flow patterns and directions, they
outcome of product design is much less predictable. Customers are essential to delight customers with personalization.
and manufacturers may be able to foresee to products that will Therefore, DFMP assumes an ecosystem thinking of product
satisfy the needs of customers within the capability of producers to innovation and service operations with the underpinning business
deliver. It could include discovery of latent customer needs based models that enable the sustainability of providers and customers
on customer profile, interactive behavior patterns, and other relationship. A product ecosystem can be specified as functional
pertinent information which reflect the cultural background, performance related design requirements with the support of
personal likings, personal tastes and aesthetics preferences. These foundation technical and business system, along with customer
are factors that can substantially differentiate from others. related interface. The premise of DFMP is the formulation of a
product ecosystem perspective of the product.
2.2. Customer experience Thus, the product ecosystem in personalization can be viewed
as three tiers, core infrastructure, hard components, and soft
The most significant extension of personalization can be characteristics. Differing from conventional design focused only on
attributed to the inclusion of experience not only in usage, but the tangible set of features, DFMP needs to take these three tiers
also buying, order processing, delivery, installation, repair, into consideration simultaneously in order to invoke active
maintenance, and disposal. Besides products’ functional perfor- customer participation and enhance customer experience.
mance, the experience of customers through these stages is a vital
part of customer satisfaction as well. The progression of the value- (1) The core infrastructure is the foundation of a product. It is
added, extending beyond physical resource intensive functional similar to the product family architecture (PFA) for a product
components to intangible customer personal feelings has been platform in mass customization. The core infrastructure goes
widely reported in many industries from automobile, appliances, beyond the technical architecture of the product ecosystems, it
computer products to telecommunication [7]. also include business models that provide fair rewards and
Customer experience is usually latent and difficult to exploit motivate contribution among all participating parties. To
through marketing analysis. It originates from evolution of the ensure the effectiveness and efficiency of ecosystems, the
user’s affective states triggered by stimuli (events) along with a stability of core infrastructure is important.
chain of cognitive tasks. The scope of design methodology (2) The hard components consist of tangible physical elements
traditionally focused on physical products. It starts with customer that can be customized through configuration and aggregation.
needs in physical space and with emphasis on functional These components are changeable to meet physical functional
performance related requirements. Customization is in no excep- requirements.
tion. A common example in customization, configuring computers (3) The soft characteristics address customer experience from
mostly deals with selection of disk drives, CPU power, memory, human–product–ambience interactions. They may include not
and other peripherals. It seldom involves the affective and only the software part of products, but also user interaction,
cognitive factors that constitute the interactions with ambience, service process, aesthetics outlook, touch and feel of products.
particularly in cases that involve a collection of products (objects) Soft characteristics are often easier to modify and thus easy to
and human users. On the other hand, personalization has to deal enhance the value for personalization.
with the affective, cognitive and functional requirements con-
currently. In order to ensure positive customer experience, 3.2. DFMP technical framework
personalization has to go beyond the confines of physical products;
it should also include intangible aspects such as services, aesthetics DFMP calls for extensive coordination among many disciplinary
and other cognitive aspects that induce the total experience. and functional groups. In general, it goes beyond the traditional
Customer experience realization is based on product’s functional territory of product design and involves more front-end business
performance, so that cognitive and affective needs elicitations can and marketing issues. At the backend, the DF’X’ mindset is
also lead to identification of unexpected performance related extended to take into account downstream production and
requirements. Desired customer experience can be further logistics concerns in the front-end design. Such a complex
improved by satisfying those unexpected requirements. Thus, engineered system requires systematic management throughout
functional performance related design requirements and customer design decision making. Fig. 1 illustrates the decision framework of
experience related requirements are positively correlated. DFMP along the entire spectrum of product realization. In a holistic
M.M. Tseng et al. / CIRP Annals - Manufacturing Technology 59 (2010) 175–178 177

Fig. 1. DFMP technical framework.

view, DFMP encompasses five domains, namely the customer, implementing value chain platforms into online personalization
functional, physical, process and logistics domains. Design for engines that can provide recommendations on latent CNs [10].
personalization involves a series of ‘‘what-how’’ mappings
between these domains. 3.2.3. Process and service platforms
The back-end issues associated with personalization involve
3.2.1. Product planning the process and logistics domains, which are characterized by
The customer domain is characterized by a set of customer process variables (PVs) and logistics variables (LVs), respectively.
needs (CNs) that are generally categorized as known CNs, which The mappings from DPs to PVs and to LVs entail process design and
are usually functional performance related customer needs and supply chain planning. The main concern of process design is to
latent CNs, which are difficult to obtain through marketing take advantage of existing capabilities and utilize repetitions in
analysis. Latent CNs include user experience related needs and production planning. The process view of personalization is
unknown performance related needs. The CNs are translated into enacted as service delivery processes. Identification of changeable,
functional requirements (FRs) in the functional domain, in which adaptable and reconfigurable processes and formulation process
the specification of a product ecosystem is formulated by taking platforms are deemed to be the fundamental issues of process
into account customer and engineering concerns. Hard and soft FRs reuse [11]. Likewise, in the logistics domain, the economic
are identified with available core infrastructure, product technol- fulfilment of personalization relies on changeable, adaptable and
ogies and marketing support. Hard FRs are designed to be reconfigurable supply and delivery networks [3].
customizable through configuration, whereas soft FRs are to be
fulfilled by personalization of customer-unique value chains based 3.3. Active customer participation
on changeable, adaptable and reconfigurable service and delivery
processes. Customization suffices with a process of customers’ making
The mapping between the customer and functional domains choices passively from pre-defined offerings. Based on these
constitutes the front-end issues associated with personalization. choices, firms can then gear up supply chains, production process
Such a planning task always starts with existing product portfolio to fulfil the orders with little or even no customer participations. In
and conforms to those common practices of order configuration cases where customer choices are based on a set of objective
and sales force automation. The exploration of soft FRs involves parameters, for example, engineering machine designs, customer
intensive customer interactions. Customer participations are participations increase little value. However, for personalization,
necessary to elicit latent customer needs. User innovation toolkits, active customer participation in the design process impacts not
data mining and Web learning lend themselves to be the main only in eliciting innate requirements but also total quality of a
techniques of customer requirement acquisition and reasoning product.
about user experience [8,9]. Furthermore, customer co-creation process and design of
human–product interactions, two major design tools commonly
3.2.2. Product and value chain platforms used in personalization, often require active customer participa-
Personalization solutions are generated in the physical domain tion. Obviously, active customer participation is also crucial to
by mapping FRs to design parameters (DPs) based on the shared satisfy user experience related requirements, because experience
product and value chain platforms. The fulfillment of hard FRs is created through a chain of human cognitive activities. Therefore,
involves typical decisions regarding product family design and active customer participation is an important design driver for the
configuration. For personalization of soft characteristics, custo- whole process, which directly influences the final product offering
mer-unique value chains must be designed in such a way that in personalization.
customer participation within a product ecosystem can be
separated into a series of value-generating activities, in which 4. Illustrative example
customer experience is recorded and analyzed. Usability study is
always useful to design changeable and adaptable work flows that Living space design accordance with its end users’ personal
enable customer co-creation and accommodate development of tastes and preferences presents a typical case of personalization. A
latent customer needs. Also concerned should be cost advantages living room involves customers’ affective, cognitive and functional
of personalized value chains. Similar to the wisdom of reusing requirements, along with many objects inside the room. Custo-
proven design elements, formulating common value chain plat- mers’ interacting with the room ambience constitutes a product
forms is deemed to be an effective means to achieve mass ecosystem. In this design case, hard FRs refer to the technical
production efficiency. New cyber-physical platforms, such as Web specifications of room design like square-feet of the room and
2.0, cloud computing, P2P and SecondLife, offer great potential for types of the objects that a particular customer prefers to. The hard
178 M.M. Tseng et al. / CIRP Annals - Manufacturing Technology 59 (2010) 175–178

Fig. 2. A living room design ecosystem.

FRs can be easily customized using a configurator, which basically be further analyzed and optimized in terms of shared-surplus
visualizes various combinations of pre-defined DPs and prompts formulation consisting of both customer satisfaction and producer
the customer to select one of them. For example, there are quite a capacity.
few space planning software available from stores like Home Depot
and Lowe’s. 5. Conclusion
What really deserves personalization is regarding those soft FRs
related to customer perception (affective needs, e.g., feel cosy) and Personalization is envisioned as one of the major drivers for the
serviceability (cognitive tasks, e.g., entertain guests), which next transformation of the global economy. As opposed to
require customers’ participations to identify. Fig. 2 shows a customization which emphasizes on design for market segments,
screenshot of a prototype virtual personalization system devel- personalization considers each customer as an individual and
oped in SecondLife [12]. The customer is represented by an avatar explores customer’s innate preference and implicit needs. It often
who is invited to ‘walk through’ different room plans and change yields a holistic solution with product and experience that are
the design on his own using the toolkit built into the virtual better than their expectation within budget and schedule. Design is
environment. The toolkit is an open source library of living room the key to realize mass personalization. Successful DFMP includes
objects, which keeps updated and accumulated from user inputs. product ecosystem as kernel, a technical design framework as the
For each testing scenario, the objects in the room ambience are design context and customer co-creation to drive the process.
tagged, such that the avatar’s interaction with each object during
the walking through can be recorded and tracked for the References
personalization engine to analyze and extract the patterns of
usability and preferences, and in turn to recommend personalized [1] Gilmore JH, Pine II JB. (2000) Markets of One. Harvard Business School Press,
Boston, MA.
room design. In this case, a foldable sleeper sofa is recommended
[2] Tseng M, Jiao J (1996) Design for Mass Customization. CIRP Annals-Manufac-
for his design, as the engine manages to reveal his latent need for turing Technology 45(1):153–156.
accommodating his guests in the living room. The customer thus [3] Jiao J, Simpson TW, Siddique Z (2007) Product Family Design and Platform-
becomes delighted as design can adapt to his unexpected need. based Product Development: A State-of-the-art Review. Journal of Intelligent
Manufacturing 18(1):5–29.
In this case, the key of personalization is context awareness, [4] Kasanoff B (2009) The Personal Economy. Proceedings of the 5th World Con-
namely relevance to customer latent needs. At the backend of the ference on Mass Customization and Personalization, Helsinki, Finaland.
prototype system, a case-driven personalization engine is devel- [5] Kumar A (2007) From Mass Customization to Mass Personalization: A Strategic
Transformation. International Journal of Flexible Manufacturing Systems
oped based on the case-based reasoning paradigm. Recommenda- 19(4):625–629.
tions are composed with reference to a priori case examples. The [6] Kano N, Seraku K, Takahashi F, Tsuji S (1994) Attractive Quality and Must-be
rough set theory [13] is applied to formulate user activity models Quality. The Journal of the Japanese Society for Quality Control 14(2):24–36.
[7] Pine II JB, Gilmore JH (1999) The Experience Economy. Harvard Business School
by recognizing typical living room activities from recorded Press, Boston, MA.
histories of avatars’ interactions with the tagged objects embedded [8] Tseng MM, Kjellberg T, Lu SCY (2003) Design in the New e-Commerce Era. CIRP
in the virtual environment. The K-optimal rule discovery technique Annals-Manufacturing Technology 52(2):509–519.
[9] Seliger G (2001) Product Innovation—Industrial Approach. Annals of the CIRP
is employed to identify underlying mapping relationships from 50(2):425–443.
soft FRs to design elements. Based on conjoint analysis of soft FRs, a [10] Jiao RJ, Xu Q, Du J, Zhang Y, Helander M, Khalid HM, Helo P, Ni C (2007)
weighted ordinal logistic regression model is developed as a Analytical Affective Design with Ambient Intelligence for Mass Customization
and Personalization. International Journal of Flexible Manufacturing Systems
benchmark to evaluate the quality of each recommendation. For
19(4):570–595.
example, Fig. 2 also shows a result of evaluating association rules [11] Wiendahl HP, ElMaraghy HA, Nyhuis P, Zäh MF, Wiendahl HH, Duffie N, Brieke
for latent need ‘feel up-scaled’, in which seat material is found as M (2007) Changeable Manufacturing—Classification, Design and Operation.
the most significant factor for this particular association according CIRP Annals-Manufacturing Technology 56(2):783–809.
[12] Secondlife, www.secondlife.com.
to the significance value of weighted ordinal logistic regression. [13] Abraham A, Falcón R, Bello R (2009) Rough Set Theory: A True Landmark in Data
Before making a design decision, this recommendation still need to Analysis. Springer, Berlin.

You might also like