100

Int. J Sup. Chain. Mgt Vol. 3, No. 4, December 2014

An Upgradable Product Design Method for

Improving Performance, CO2 Savings, and
Production Cost Reduction: Vacuum Cleaner
Case Study
Masato Inoue#1, Shuho Yamada#2, Tetsuo Yamada*3, Stefan Bracke**4
#
Department of Mechanical Engineering Informatics, Meiji University
1-1-1 Higashi-Mita, Tama-ku, Kawasaki, Kanagawa 214-8571, Japan
1
m_inoue@meiji.ac.jp
2
ce42088@meiji.ac.jp
*
Department of Informatics, The University of Electro-Communications (UEC Tokyo)
1-5-1 Chofugaoka, Chofu-shi, Tokyo 182-8585, Japan
3
tyamada@uec.ac.jp
**
Chair of Safety Engineering and Risk Management, University of Wuppertal (BUW)
Gaußstraße 20, Wuppertal 42119, Germany
4
bracke@uni-wuppertal.de

Abstract—Customers often discard products without production cost of the supply chain and quality is
considering the environmental load, because of the established [2]. Therefore, an environmentally
deterioration of their value even though they are fully
functional. An upgradable product design can
conscious product design is essential. Design
enhance the product value and stop the entire methods such as the 3R, which refers to reuse,
product replacement and disposal by replacing only reduce, and recycle, and upgradable product design
the components responsible for the decrease in value. methods have been studied [3]. An upgradable
This paper proposes an upgradable product design
method for improving product performance, product design method is inherent to the disposal of
incurring CO2 savings, and production cost reduction products. A sustainable society needs to consider
while increasing the product value and extending the product upgradability as a new sort of “evolution”
value lifespan by exchanging components closely
[4] while harmonizing environmental and financial
related to the deterioration in value. In addition, this
paper proposes a method that can specify future demands in designing the supply chain [5].
product performance, effective upgradable product Ishigami et al. [6] proposed a method for upgrade
components, and the effect of the upgrade on other planning based on the prediction of customer
product components. Finally, this paper discusses the
applicability of the proposed method by considering demands. They specified a line-up of candidate
the designing of a vacuum cleaner and customer upgradable components for each generation.
demands such as performance, noise, and energy Furthermore, Fukushige et al. [7] developed a
savings. prototype system for product upgradability based
Keywords—Sustainability, upgradability, CO2 savings,
on 3D-CAD. However, these methods focus on the
cost reduction, customer demand, set-based design
physical product performance without quantitative
consideration of requirements such as production
1. Introduction
cost and environmental load (CO2 savings). Thus,
A sustainable society requires changes in the this paper focuses on an upgradable product design
traditional paradigm of mass production and method for improving product performance and
consumption. Companies and countries need to CO2 savings while reducing the production cost.
reduce their environmental loads [1]—especially at This method also considers future unknown
the product design phase, where at least 80% of the product requirements and design information. We
assume that customers discard a product when its
perceived value has decreased with time below the
______________________________________________________________
International Journal of Supply Chain Management perceived market value of new products. In
IJSCM, ISSN: 2050-7399 (Online), 2051-3771 (Print) addition, this paper proposes a method for
Copyright © ExcelingTech Pub, UK (http://excelingtech.co.uk/)
specifying future product performance and
functions, effective upgradable product
 101
Int. J Sup. Chain. Mgt Vol. 3, No. 4, December 2014

components, and the effect of the upgrade on other

product components. Finally, this paper discusses Upgrade planning
the applicability of the proposed upgradable

Determination of the upgrade product
product design method by considering a vacuum

Definition of the upgrade time
cleaner and customer demands such as

Creation of the product and the
performance, noise, and energy savings. component database

Creation of the product and the
2. Upgradable Product Design component roadmaps
Method
2.1 Purpose and methodology
Developing function network diagram
An upgradable product design method seeks to
design products that are capable of adaptation to
Representation of the input–output
future product performance enhancement and relations between performance
functions by predicting the product performance criteria and product components
and functions that will be required at the time of the
Definition of positive and negative
upgrade. There are two basic implementations of parameters of performance criteria
the method:

(a) upgrade by exchanging components and

(b) upgrade by adding components or Upgradable performance
modifying the product structure. criteria and components

In this study, we focus on the first kind. Because
Investigation of customer demands by
future enhancements of product performance and application of QFD
functions need to be predicted, the proposed
Definition of potential upgradable
method must consider fuzzy design information. components
The traditional design practice often considers the
engineering design as an iterative process; that is, it
quickly develops a “point solution,” evaluates it on Upgrade-affected performance
the basis of multiobjective criteria, and then criteria and components
iteratively moves on until it reaches a satisfactory
point solution. However, there is no theoretical
Definition of upgrade affected
guarantee that the process will ever converge and components
produce an optimal solution. In addition, the use of
Definition of upgrade affected
a point solution does not provide information about performance criteria
uncertainty. In this study, the uncertainty in the
design information is represented as a range of
values, and we use a preference set-based design
(PSD) method [8], [9] to estimate future product Application of the PSD method and
performance enhancements and upgrade features. evaluation of the solution set
The PSD method results in a range of design
solutions. Figure 1 shows the proposed
Definition of the functional
methodology. requirements and ranged design
variables
In this study, we define two product lifetimes:
Determination of the relationships
durable life and value life [10]. Durable life between product performance and
measures the duration over which the failure rate of component design variable
a product or a component remains below a defined
threshold. Conversely, value life represents the
Figure 1. Procedure of the proposed upgradable
duration over which the product value as perceived
product design method
by the customer remains above a defined threshold.
Products such as personal computers and
our methodology is to reduce the extent of product
smartphones are discarded even though they are
disposal and the resulting environmental load by
fully functional. In this case, value life is shorter
increasing the product value life by exchanging
than durable life. Therefore, the primary purpose of
components with relatively short value life or by
 102
Int. J Sup. Chain. Mgt Vol. 3, No. 4, December 2014

adding new components in accordance with the diagram

above description.
A function network diagram shows the input and
2.2 Upgrade planning output relation between performance criteria and
product components. This diagram is used to
We establish the criterion of the upgrade time. The analyze the upgrade components. Figure 2 gives an
definition of the upgrade time is based on several example of a function network diagram for a
factors such as product upgrade cycle, disposal vacuum cleaner. In this diagram, performance
cycle, or administrative strategy. In addition, criteria and product components are represented by
product and component databases are created. individual graphics. Positive parameters denote
These databases contain manufacturers, model conditions where a higher or larger value represents
numbers, launch times, and product and component better performance, while negative parameters
performance, or design variables such as storage denote conditions where a lower or smaller value
capacity, weight, and dimensions. Based on these represents better performance. The input–output
databases, we create product and component relations between performance criteria and product
roadmaps to evaluate the temporal distribution of components are connected by straight lines, and on
performance criteria and design variable values. these lines, the relevant design variables are
The upgrade time and product and component described. Therefore, designers can easily search
roadmaps are used for configuring the performance for components that are related to upgrade
requirements of the products and components. performance by following the input–output lines.
When a product has not yet been launched,
roadmaps of similar products can be used for In the case of a vacuum cleaner (Figure 2), for
market predictions and for configuring the example, suction power is the chosen target for
performance requirements. upgrade performance. The components affecting
the performance are the motor, turbine fan, body,
2.3 Developing the function network power unit, and are candidates for upgrade.

Hp
Power unit Waste heat

N Suction
Motor
power F

T Operation
noise S
Turbine fan
CO2 emission D
(Use phase)

Energy
consumption E

Vi Penetration rate of a filter Exhaust

Body Filter
cleanness

Hp : Waste heat of power unit Dt : Diameter of turbine fan Positive

Hm : Waste heat of motor Cf : Coefficient of noise (turbine) parameter
Ep : Electric power Cm: Coefficient of noise (motor)
Vi : Inner volume of body a : Energy conversion efficiency
T : Torque (motor) f : Coefficient of air flow resistance Negative
N : Revolution (motor) ht : Pressure of inside of the body parameter

Figure 2. Function network diagram for a vacuum cleaner and graphical representation
of the performance criteria and product components
 103
Int. J Sup. Chain. Mgt Vol. 3, No. 4, December 2014

2.4 Upgradable performance criteria of the required product performance and functions,
and components as well as the design variables of the components,
are needed. The equations show the relations
Upgradable performance criteria are defined as the between product performance and functions and
product performance upon which customer component design variables. In the absence of
emphasis is placed and evaluated by the application equations, the designer should define approximate
of quality function deployment (QFD) [11], or by equations based on the performance parameters and
performance criteria with short value life. Using the design variables in the product and component
function network diagram, a designer searches for databases.
components that have close relations with the
chosen upgrade performance criterion and defines The range of the required product performance
and functions and the designer configurable range
them as potential upgradable components. When
the number of upgradable components is identified, of design variables are configured in accordance
the candidates are narrowed down by considering with the distributions in the product and component
roadmaps. A conclusive point-based design
the balance between upgrade performance criteria
and possible upgrade-affected performance criteria proposal is selected from a set of design solutions
and components, as described below. and a preference number. When the design
proposal must be modified, the designer should
2.5 Upgrade-affected performance search for a design proposal that satisfies the
modified requirements from the set of design
criteria and components solutions. However, in the absence of a design
proposal in the set of design solutions, the designer
Upgrade-affected performance criteria indicate that
should redefine the required performance and
performance changes by exchanging upgrade
design variables and apply them to the PSD method.
components. In addition, we define the component
that is closely related with an upgrade-affected
performance criterion as the upgrade-affected
3. Case Study: The Design of a
component. Upgrade-affected performance criteria Vacuum Cleaner
are similarly identified as the upgrade components
by using the function network diagram. For In this study, we discuss the application of the
example, a designer defines the performance of proposed method to designing a vacuum cleaner.
suction power as the upgrade performance criterion According to the trade-up cycle to a new model, we
and the turbine fan component as the upgrade assume an upgrade time of approximately seven
component. In this case, the waste heat and years (for a customer trade-up percentage of 60%)
operation noise are defined as upgrade-affected from the launch time of the first-generation product.
performance criteria because the upgrade of the To understand the trend of the performance
turbine fan causes the diameter Dt of the turbine fan requirements and design variables, we create
and the noise coefficient Cf of the turbine to databases for launched products and performances.
increase. Therefore, the power unit and motor are Figure 3 shows the product databases for the
upgrade-affected components. Possible approaches suction power of vacuum cleaners manufactured by
for mitigating this condition can be developed, for three companies (A, B, and C) from 2005 to 2013.
example, by building low-level waste heat and low- The component database includes motors
noise motors into a first-generation vacuum cleaner, manufactured by a single company in 2013 because
or developing and upgrading a motor with high there are no data for motors manufactured before
suction power, low-level waste heat, and low noise 2012.
while upgrading and changing the motor with the
turbine fan. These approaches are narrowed down Figure 2 shows the function network diagram of
in the same way as that of the upgrade components. a vacuum cleaner. Using the QFD result, we define
suction power F [W] as the upgrade performance
2.6 Application of the PSD method and with high level of value degradation. Using the
information in Figure 2, the motor is configured as
evaluation of the solution set
an upgrade component. In addition, the
performance criteria affected by the upgrade are
In this study, we apply the PSD method to the
operation noise S [dB], amount of CO2 emissions D
proposed method to obtain the range of required
[g] in the use phase, energy consumption E [W] in
product performance and functions and the range of
the use phase, and total production cost. The
the component design variables that can realize this
product performance requirements and the range of
performance and function range. To obtain such
design variables based on the product database and
range of design solutions, the equations and range
 104
Int. J Sup. Chain. Mgt Vol. 3, No. 4, December 2014

[scenario 3] the customer requires low-energy

consumption.

4. Results and Discussion

With respect to the relations among suction power
(upgraded performance), operation noise (upgrade-
affected performance criterion), CO2 emission
(upgrade-affected performance), and production
cost, Figures 4–7 show the performance-oriented
solutions (scenario 1: performance-oriented
customer), noise avoidance solutions (scenario 2:
silence-oriented customer), and ecology-driven
solutions (scenario 3: energy-savings-oriented
customer).

We obtain the set of design solutions that include

Figure 3. Databases for launched products and
performance criteria, production cost, and
performance of vacuum cleaners
environmental load (CO2 emission) while
manufactured by three companies
considering future fuzzy design information.
roadmap are configured. Finally, the set of design
Figure 4 compares suction power according to
solutions is calculated using the equations between
customer demands. The results suggest that suction
product performance and design variables from the
power is upgraded in all scenarios. In particular, the
PSD method.
performance of the second-generation product for
The range of design variables can increase and scenario 1 is upgraded by about 15% compared
decrease relative to the reference values that are with that of the first-generation product.
assumed as the design variables of the first-
Figure 5 compares operation noise according to
generation product. The total production cost for
customer demands. Operation noise is the
the upgraded product has multiple relations with
performance criterion affected by the upgrade.
the costs of the upgrade components and the
Therefore, the operation noise in the second-
affected components (i.e., the motor and turbine
generation product for scenario 1 is greater by
fan). We assume that the cost of the components
about 10% compared with that in the first-
increases relative to the difference between the
generation product. However, the operation noise
design variables of the first (reference value) and
in the second-generation product for scenarios 2
second generations. Therefore, the component cost
and 3 is reduced by about 6% and 3%, respectively.
(excluding the motor and turbine fan) is higher than
the cost of the nonupgraded components. Figure 6 compares CO2 emission according to
Presumably, if the range of design variables customer demands. CO2 emission is also an
decreases, then the increase in total production cost upgrade-affected performance criterion. The results
is less than that in the case with increasing range of suggest that CO2 emission is reduced in all
design variables. scenarios. In particular, the CO2 emission in the
second-generation product for scenario 3 is reduced
We assume that the product performance and
by about 23% compared with that in the first-
design variables of the components of the second-
generation product.
generation product without upgrade are equal to the
upgraded product parameters. In addition, we Figure 7 compares production cost according to
assume that the cost of the second-generation customer demands. The results suggest that the
product without upgrade is equal to that of the first- total production cost of the upgraded product is
generation product without upgrade. reduced by about 25% compared with that of the
nonupgraded product, and that suction power
We group customer demands into three
(upgraded performance) satisfies the required range.
scenarios:
Therefore, we conclude that the proposed method
[scenario 1] the customer requires high suction can obtain a set of product design solutions of
power, performance criteria, environmental load, and
[scenario 2] the customer requires a low-noise production cost according to the various customer
vacuum cleaner, and preferences. However, the designers should define
 105
Int. J Sup. Chain. Mgt Vol. 3, No. 4, December 2014

first-generation product first-generation product

[11.3, 15.8] (%) [-5.5, -0.3] (%)

[Scenario 1] Performance
[Scenario 1] Performance
高性能重視
高性能重視
-oriented customer
-oriented customer
[0.5, 1.5] (%) [-4.3, -3.0] (%)
[Scenario 2] Silence
[Scenario 2] Silence
静音性重視
静音性重視
-oriented customer
-oriented customer
[2.7, 7.0] (%) [-23.7, -19.5] (%)
[Scenario 3] Energy- [Scenario 3] Energy-
省エネ重視
省エネ重視
savings-oriented customer savings-oriented customer

600 650 700 400 500 600 700

Suction power [W] CO2 Emission [g]

Set of design solutions by PSD method Set of design solutions by PSD method

Required performances Required performances

Figure 4. Comparison of suction power Figure 6. Comparison of CO2 emissions

according to customer demands according to customer demands

first-generation product Cost for second-

generation product
Cost for first-
[7.6, 10.0] (%) generation product
[Scenario 1] Performance
高性能重視
-oriented customer Nonupgraded
NO UG
[-6.6, -5.9] (%) product
[Scenario 2] Silence
静音性重視
[22.4, 27.6] (%) [-25.8, -22.6] (%)
-oriented customer [Scenario 1]
[-3.4, -0.7] (%) 高性能重視
Performance
[Scenario 3] Energy-
省エネ重視
savings-oriented customer -oriented customer
[Scenario 2] [-25.5, -22.8] (%)
55 60 65 静音性重視
Silence
Operation noise [dB] -oriented customer
[Scenario 3] [-24.9, -21.6] (%)
Set of design solutions by PSD method 省エネ重視
Energy-savings-
Required performances oriented customer
0 20,000 40,000 60,000
Figure 5. Comparison of operation noise Production cost (Yen)
according to customer demands Minimum cost
for first- Minimum cost
generation product for upgrade
the rate of increase in the first-generation product
Maximum cost
costs for the upgraded product and include it in the for first-generation Maximum cost
product upgrade requirements. product for upgrade

5. Conclusions Figure 7. Comparison of production cost

according to customer demands
This paper proposed an upgradable design method
for obtaining a set of solutions that satisfies value life. We demonstrated the usefulness of this
multiple product performance criteria, CO2 savings, diagram by considering the case of a vacuum
and reduces production cost by considering cleaner. It is essential to calculate the product
uncertain design information of the various manufacturing environmental load, which we have
customer preferences. To obtain the set of solutions, not done, rather than the product environmental
the PSD method is used to propose an increased load in use.
product value and extended product life by
exchanging components with values below 6. Future Work
specified thresholds. In addition, we proposed the
function network diagram to define the product In this study, the set of requirements that include
performance criteria and components with short future uncertainty is arbitrarily predicted on the
 106
Int. J Sup. Chain. Mgt Vol. 3, No. 4, December 2014

basis of the distribution of the product roadmap. cycle engineering, Elsevier, Norway, pp. 379–
Moreover, this prediction is based on the designer’s 384, 2014.
knowledge and experience. Therefore, predicting [5] Yamada T., “Design of closed-loop and low-
the set of requirements is a subject of future study. carbon supply chains for sustainability”,
In the case of a vacuum cleaner, the motor is Manufacturing and Environmental
upgraded to upgrade suction power. However, we Management, National Political Publishing
achieved performance upgrade just by upgrading House, pp. 211–221, 2012.
the turbine fan. Therefore, we need to propose the [6] Ishigami Y., Yagi H., Kondoh S., Umeda Y.,
most suitable quantitative method for exchanging Shimomura Y., Yoshioka M., “Development
of a design methodology for upgradability
components by understanding the effect of
involving changes of functions”, 3rd
exchanging components. In general, durable life is
international symposium on environmentally
longer than value life. Therefore, the component
conscious design and inverse manufacturing
upgrade should occur before durable life ends and (EcoDesign 2003), IEEE, Japan, pp. 235–242,
at about the end of value life. Nonetheless, a 2003.
method is required to define the optimal balance [7] Fukushige S., Arino M., Umeda Y.,
between value life and durable life for designing “Computer-aided design for product
components with sufficient durable life. upgradability under geometric constraints”,
7th international symposium on
Acknowledgments environmentally conscious design and inverse
manufacturing (EcoDesign 2011), Springer,
This work was partially supported by a Research Japan, pp. 828–831, 2011.
Project Grant by the Institute of Science and [8] Inoue M., Nahm Y.-E., Okawa S., Ishikawa
Technology, Meiji University, and JSPS H., “Design support system by combination of
KAKENHI Grant Number 26870628. 3D-CAD and CAE with preference set-based
design method”, Concurrent Engineering:
References Research and Applications, Vol. 18, No. 1,
pp. 41–53, 2010.
[1] Kyoto Protocol Target Achievement Plan, [9] Inoue M., Nahm Y.-E., Tanaka K., Ishikawa
Ministry of the Environment of Japan, H., “Collaborative engineering among
http://www.kyomecha.org/document/pdf/kp_a designers with different preferences:
chieveplan.pdf, (14-08-2014). application of the preference set-based design
[2] Singhry H. B., Rahman A. A., Imm N. S., to the design problem of an automotive front-
“The potential moderating role of supply side frame”, Concurrent Engineering:
chain capabilities on the relationship between Research and Applications, Vol. 21, No. 4,
supply chain technology and concurrent pp. 252–267, 2013.
engineering in product design”, International [10] Bracke S., Michalski J., Inoue M., Yamada T.,
Journal of Supply Chain Management, Vol. 3, “CDMF-RELSUS concept: reliable products
No. 2, pp. 132–139, 2014. are sustainable products- influences on
[3] Kobayashi H., “Product lifecycle planning- A product design, manufacturing and use
practice for ISO/TR14062”, Ohmsha, in phase”, International Journal of Sustainable
Japanese, 2003. Manufacturing, Vol. 3, No. 1, pp. 57–73, 2014.
[4] Pialot O., Millet D., “Why upgradability [11] Akao Y., “QFD: quality function deployment-
should be considered for rationalizing integrating customer requirements into
materials?” 21st CIRP conference on life product design”, Productivity Press, 2004.