Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 61 (2017) 245 – 250

The 24th CIRP Conference on Life Cycle Engineering

Interlinking multiple decision variables over different life cycle stages to

realize effective reuse and recycling from a strategic viewpoint
Shinsuke Kondoha*, Hitoshi Komotoa, Keijiro Masuia
a
National Institute of Advanced Industrial Science and Technology (AIST), 1-2-1, Namiki, Tsukuba, Ibaraki, 305-8564, Japan

* Corresponding author. Tel.: +81-29-861-7067; fax: +81-29-861-7067. E-mail address: kondou-shinsuke@aist.go.jp

Abstract

Effective reuse and recycling of End-Of-Life (EOL) products is essential for improving resource efficiency. Many studies have been conducted
for enabling effective reuse and recycling of EOL products. However, their individual application is not sufficient for the purposes because the
main concerns dealt with in each method are strongly interrelated. The objective of the study is proposing a method for interlinking multiple
factors over different life cycle stages so that multiple stakeholders involved in different life cycle stages can collaboratively find out more
effective solutions than those individually achieved by each stakeholder from a strategic viewpoint.
©
©2017
2017TheTheAuthors. Published
Authors. by Elsevier
Published B.V. This
by Elsevier is an open access article under the CC BY-NC-ND license
B.V.
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the 24th CIRP Conference on Life Cycle Engineering.
Peer-review under responsibility of the scientific committee of the 24th CIRP Conference on Life Cycle Engineering
Keywords: "Reuse; recycling; product design; capacity planning"

1. Introduction effective solutions than those individually achieved by each

Effective reuse and recycling of End-Of-Life (EOL)
products is essential for improving resource efficiency. Many
studies have been conducted for enabling effective reuse and
recycling of End-of-Life (EOL) products [1-10]. Some focused
on the product design methods that ease disassembly and
recycling of EOL products [2, 3]. Others focused on the
planning and sequencing of disassembly operations [4, 5],
capacity planning of remanufacturing facilities [6], material
separation technologies [7], the design of reverse network for
EOL products [8], and functional sales and service offering
design [9,10]. Although these methods and tools are quite
effective for widening the opportunities for product reuse and
recycling, their individual application is not sufficient for the
purposes because the main concerns dealt with in each method
are strongly interrelated. Thus, the simultaneous consideration
of these factors is essential for solving the problem.
The objective of the study is proposing a method for
interlinking multiple factors, some of which are to be
designated as decision variables, over different life cycle stages.
Based on such interlinking, multiple stakeholders involved in
different life cycle stages can collaboratively find out more
stakeholder.
This paper especially focuses on interlinking product design
variables to those affecting profitability in reuse and recycling
processes so that the original equipment manufacturers (OEMs),
second hand product retailers, and recyclers can collaboratively
decide more optimal product structure with adequate capacity
in reuse and recycling facilities. As a first step of this attempt,
a linear programing model is introduced to quantitatively
correlate such multiple factors. Through a simplified case study
of a mobile phone recycling, the possibility of the model for
interlinking these factors is demonstrated.
The rest of the paper is organized as follows. Section 2
introduces a linear programing model that are used in the study.
Section 3 explains the method that computes the optimal reuse
and recycling plan for EOL products with its optimal
assignment to corresponding firms. Section 4 provides a
simplified calculation example of a mobile phone. Section 5
discusses the advantages and drawbacks of the proposed model.
Section 6 concludes the paper.

2212-8271 © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the 24th CIRP Conference on Life Cycle Engineering
doi:10.1016/j.procir.2016.11.256
246 Shinsuke Kondoh et al. / Procedia CIRP 61 (2017) 245 – 250

2. The model interlinking product design to reuse and EOL product, sub-assembly, component, or mixed material
recycling planning fraction) that can be sold to reuse or recycling firms. The root
item represents the product itself and a leaf item represents any
2.1. Interdependency between product design and item that is to be reused or discarded without any further
reuse/recycling of EOL products disassembly operations. A set of edges between the parent item
and its child items corresponds to the disassembly or
Among multiple criteria for evaluating the desirability of dismantling process that is required to obtain the child items
reuse/recycling of EOL products, the paper especially focuses from the parent item. Fig. 2 shows a simplified RRT of a mobile
on the economic benefit as a first step. Fig.1 shows the possible phone. A mobile phone can be disassembled into two parts:
interdependency among decisions in product design and liquid crystal display (LCD) and printed-circuit board (PCB).
reuse/recycling process planning from the economic viewpoint. LCD can be further dismantled to mixed material factions, each
According to the interview result with engineers in recycling of which contains steel, aluminum, and other materials. Rare
firms, reuse/recycling process parameters such as capacity and metal fraction (e.g., tantalum-rich fraction) can also be
location of disassembly/separation facilities and maturity of dismantled from PCB. In addition, the average market price for
disassembly/separation technology, which directly affect each item is given as a number beside each node in the RRT.
disassembly, separation, and transportation cost of EOL Negative price for “Others” implies that the fraction cannot be
products as well as the quality of the retrieved items, are sold but should be landfilled or incinerated by its producer’s
regarded as key decision variables for the reuse/recycling cost.
planning. However, the decisions made in product design (e.g.,
material composition, fastening method, and geometrical 0RELOH >\HQ@
structure etc.) have also significant potential for improving the SKRQH

profitability of the reuse and recycling. To highlight such
/&' >\HQ@ 3&% >\HQ@
potential, quantitative evaluation of the effect of product design 

is indispensable. Thus, a linear programing model which
correlate product design with profitability of reuse/recycling of 5DUH
6WHHO $OXPLQXP 2WKHUV 3&%䇺 PHWDOV
EOL product is proposed in the paper. >\HQ@ >\HQ@ >\HQ@ >\HQ@ >\HQ@

4XDOLW\
$PRXQW
6DOHV &DSDFLW\ Fig. 2. Simplified RRT of a mobile phone.
3ULFH
(FRQRPLF 'HPDQG
JDLQ 0DWXULW\RI
WHFKQRORJ\
&RVW 'LVDVVHPEO\
FRVW
/RFDWLRQ
6HSDUDWLRQ
FRVW
DŽďŝůĞƉŚŽŶĞ WĚŝƐŵĂŶƚůŝŶŐ WƌĞĨŝŶĞƌLJ
dŚĞƐĐŽƉĞĨŽƌĚĞĐŝƐŝŽŶ 7UDQVSRUWDWLRQ ZĞĐLJĐůŝŶŐĨŝƌŵƐ ĨŝƌŵƐ ĨŝƌŵƐ
ŵĂŬŝŶŐŽĨ FRVW
ƌĞƵƐĞͬƌĞĐLJĐůŝŶŐĨŝƌŵƐ
3XUFKDVH
FRVW
Dϭ Wϭ W͛ϭ
4XDOLW\ 3ULFH 6DOHV
3URGXFW
/ŶƚĞƌĚĞƉĞŶĚĞŶĐLJ GHVLJQ
'HPDQG $PRXQW DϮ WϮ W͛Ϯ
&ĂĐƚŽƌƐͬƉĂƌĂŵĞƚĞƌƐ &RVW (FRQRPLF
ŝŶƌĞƵƐĞͬƌĞĐLJĐůŝŶŐ JDLQ
dŚĞƐĐŽƉĞĨŽƌĚĞĐŝƐŝŽŶŵĂŬŝŶŐŽĨKƌŝŐŝŶĂů
ŽĨK>ƉƌŽĚƵĐƚƐ >ϭ dϭ
ƋƵŝƉŵĞŶƚDĂŶƵĨĂĐƚƵƌĞƌƐ;KDƐͿ
ϭϬϬŬŵ
ϱϬϬŬŵ >Ϯ dϮ
> ƌĞƵƐĞ ZĂƌĞŵĞƚĂů
Fig. 1. Interdependency between product design and reuse/recycling process &ŝƌŵƐ ĨŝƌŵƐ ƌĞĨŝŶĞƌLJĨŝƌŵƐ

2.2. The Reuse/Recycling Tree (RRT) model Fig. 3. Partial PRN of a mobile phone.

As the product design include a lot of information, it is not
feasible to directly correlate every decision variables with those
used in reuse/recycling planning. Thus, we encapsulate product
design information as Reuse and Recycling Tree (RRT) model
and correlate it with Provider and Receptor Network (PRN),
which represents possible assignment of reuse/recycling tasks
to firms.
The RRT represents a possible sequence of component reuse
and material recycling for a given EOL product in addition with
the average market price for each item recovered from the
product. Each node of the tree corresponds to the item (i.e.,
2.3. The Provider Receptor Network (PRN) model
The PRN represents assignment of the recycling (or reuse)
processes to firms, which are sparsely located in Japan. Each
node of the PRN corresponds to the reuse or recycling firm that
can execute one of the processes represented in the RRT. The
processing capacity, cost, and storage capacity of each firm are
also tagged to each node. Each edge of the PRN corresponds to
the transportation (and transaction) among two firms. The
origin node and the destination node of the edge are referred as
the provider and the receptor, respectively. The traveling
distance among two firms, based on which the transportation
 Shinsuke Kondoh et al. / Procedia CIRP 61 (2017) 245 – 250 247

cost is calculated, is also tagged to the edge. As multiple firms
can execute the same process represented in the RRT in general,
the PRN consists of multiple groups of nodes, each of which
corresponds to a set of firms that can execute the same process.
Fig. 3 shows a part of the PRN corresponding to the RRT given
in Fig. 2. There exist two mobile phone recycling firms (i.e.,
M1 and M2), two PCB dismantling firms (i.e., P1 and P2), two
LCD reuse firms (i.e., L1 and L2), two PCB refinery firms (i.e.,
P’1 and P’2), and two rare metal refinery firms (i.e., T1 and T2),
each of which can reuse/recycle a mobile phone, PCB, LCD,
PCB’, and rare metal fractions, respectively.
2.4. Procedure for finding profitable reuse and recycling plan
The procedure for finding out profitable reuse and recycling
plan is roughly divided into two phases as shown in Fig. 4. In
the first phase (i), a given environment of reuse/recycling
network with product structure are interpreted as RRT and PRN
models. Parameters that can be changed through redesign of
product structure and reuse/recycling networks (e.g., capacity,
demand and operation cost of reuse/recycling facilities,
transportation cost among firms, disassembly/separation cost of
a given product, and so forth) are treated as constants in this
phase. By using these constant values, sub optimal reuse and
recycling plan of a given EOL product is determined at first.
Then, the optimal assignment of its corresponding processes to
firms are computed by linear programming algorithm. Through
a sensitivity analysis of objective function, most influential
RRT and PRN parameters (i.e., bottleneck parameters) among
these constants are identified.
In the second phase (ii), product and process design
parameters that can control the bottleneck parameters identified
in the first phase (e.g., material composition in each component,
weights of each component, purity of recovered materials etc.)
are to be sought out and possible improvement for the
bottleneck parameters is estimated by correlating them with the
bottleneck parameters. Based on the estimated value for the
bottleneck parameters, the calculation of phase (i) is conducted
transportation and disassembly cost differ from firm to firm, the
market price for each item is used for this calculation.
From the root node in the RRT, profitability of its reuse (or
recycling) of components (or material) is calculated one by one.
First, for each item identified in the RRT, the total sales of its
child items (i.e., the items that can be retrieved from the parent
item) is calculated. Then, the sales of the child items is
compared with the price of the parent item. If the sales of the
child items is larger than the price of the parent item, then the
parent item is disassembled or dismantled, else the parent item
is to be directly reused for other purposes and is regarded as a
leaf item.
Step 3: Assignment of reuse/recycling processes
The designer selects the most profitable firms for each
process in the reuse and recycling plan determined in step 2,
considering the differences in transportation and processing
cost caused by the differences in locations, capacity, available
technologies, and inventory level of each firm in a given period.
Step 4: Calculation of objective function
Based on the result of step 3, total profit for the material
recycling (or component reuse) is calculated, as well as the
amount of items that are discarded or stored in this period.
Step 5: What-if analysis for generating design ideas
Through sensitivity analysis of objective function, most
influential RRT and PRN parameters are identified. Conducting
what-if analysis considering the variations in these parameters,
redesign ideas of products and processes are generated.
3URFHVVGHVLJQ L 0RVWLQIOXHQWLDO
557PRGHO
SDUDPHWHUV SDUDPHWHUV 557351
SDUDPHWHUVDUH
LGHQWLILHGWKURXJK
VHQVLWLYLW\DQDO\VLV
LL 3URGXFWSURFHVVGHVLJQ
SDUDPHWHUVWKDWFDQFRQWUROWKH
NH\557531SDUDPHWHUVDUH
LGHQWLILHG
*RDOLQGLFDWRUV

again.
In this way, interdependency among product structure, 351PRGHO
disassembly and recycling technologies, and capacity of 3URGXFWGHVLJQ SDUDPHWHUV
reuse/recycling firms is taken into account in finding optimal SDUDPHWHUV
3URSRVHGPRGHO
reuse/recycling plan. More detail procedure is summarized as
follows:
Step 1: Development of the linear programming model Fig. 4.Procedure for finding optimal reuse/recycling plan as well as product
The designer develops the RRT of the target product by design
identifying its possible set of components, sub-assemblies, and
mixed material fractions that can be sold in addition with the
disassembly processes required to obtain them. For the sake of
simplicity, the paper assumes that each product has only one WƵƌĐŚĂƐĞŽĨŝƚĞŵŵ ^ĞůůŝŶŐĂĨƚĞƌ
RRT. Market price for each item is also estimated. ƌĞĐŽǀĞƌĞĚĨƌŽŵ ŝƐĂƐƐĞŵďůLJͬƐĞƉĂƌĂƚŝŽŶ
K>ƉƌŽĚƵĐƚƐ proct,m,i
Then, the designer develops the corresponding PRN gt,m,i ĂƌƌLJŝŶŐŽǀĞƌƚŽƚŚĞŶĞdžƚ
identifying every available firms for each process in the RRT. ƉĞƌŝŽĚ;ƐƚŽƌĂŐĞͿ
The process cost, processing capacity, and storage capacity of /ƚĞŵƐĐĂƌƌŝĞĚŽǀĞƌ st,m,i
WƌŽǀŝĚĞƌŝ
each firm is also estimated in addition with traveling distance ĨƌŽŵƚŚĞƉƌĞǀŝŽƵƐ
ƉĞƌŝŽĚ ŝƐĐĂƌĚŝŶŐ
among firms. st-1,m,i wt,m,i
Step 2: Identification of a profitable reuse and recycling plan
Based on the RRT of the target product, a profitable reuse Fig. 5 .Input and output for a reuse/recycling firm
and recycling plan is identified in a given period. Because the
248 Shinsuke Kondoh et al. / Procedia CIRP 61 (2017) 245 – 250
3. Optimization method
Input and output for each firm in PRN model is given as
shown in Fig. 5. From the viewpoint of each firm in PRN model,
the optimization problem is described as follows: for a given
item m (i.e., items to be sold except leaf item) in the RRT with
a given amount in a given period t, the problem is to determine
the optimal amount of its child items to be sold to their
corresponding firms while minimizing the amount of
discarding and storing (i.e., carrying over to the next period).
The overall assignment of each process in the RRT is computed
by solving the problem one by one from its root item toward
leaf items. To model the problem, the following notations are
used.
3.1. Parameters
I: a set of providers’ indices;
J: a set of indices for the receptors;
Lm: a set of indices for the items recovered from the item m;
gt,m,i: the amount of item m that are purchased or generated by
provider i in period t;
pct,m,i: purchase cost of item m by provider i in period t;
spt,l,j: sales price of item l for receptor j in period t;
di,j: distance between provider i and receptor j;
tcl: unit transportation cost of item l;
pct,m,i: unit disassembly or dismantling cost of item m at
provider i in period t;
sct,m,i: unit storage cost for item m at provider i in period t;
wct,m,i: unit disposal cost for item m at provider i in period t;
pcapt,m,i: maximum processing capacity for item m at provider i
in period t;
pcapt,l,j: maximum processing capacity for item l at receptor j in
period t;
scapt,m,i: maximum storage capacity for item m at provider i in
period t.
3.2. Decision variables
xt,l,i,j: the amount of item l that are sold by provider i to receptor
j in period t;
proct,m,i: the amount of item m that are processed at i in in period
t;
st,m,i: the amount of item m that are stored by provider i in period
t;
wt,m,i: the amount of item m that are discarded as waste by
provider i in period t.
3.3. Problem formulation
The profit of processing a given item m in period t is given
as following equation:
݂ ൌ ෍ ෍ ෍ ‫ݔ‬௧ǡ௟ǡ௜ǡ௝ ൫‫݌ݏ‬௧ǡ௟ǡ௝ െ ݀௜ǡ௝ ‫ܿݐ‬௟ ൯ െ ෍ ‫ܿ݋ݎ݌‬௧ǡ௠ǡ௜ ‫ܿ݌‬௧ǡ௠ǡ௜
௟‫א‬௅௠ ௜‫א‬ூ ௝‫א‬௃ ௜‫א‬ூ
െ ෍ ‫ݏ‬௧ǡ௠ǡ௜ ‫ܿݏ‬௧ǡ௠ǡ௜ െ ෍ ‫ݓ‬௧ǡ௠ǡ௜ ‫ܿݓ‬௧ǡ௠ǡ௜ െ ෍ ݃௧ǡ௠ǡ௜ ‫ܿ݌‬௧ǡ௠ǡ௜
௜‫א‬ூ ௜‫א‬ூ ௜‫א‬ூ
(1)
The problem is formulated as maximizing f subjective to
‫݅׊‬ǡ ‫ݐ‬ൣ݃௧ǡ௠ǡ௜ ൅ ‫ݏ‬௧ିଵǡ௠ǡ௜ ൌ ‫ܿ݋ݎ݌‬௧ǡ௠ǡ௜ ൅ ‫ݏ‬௧ǡ௠ǡ௜ ൅ ‫ݓ‬௧ǡ௠ǡ௜ ൧㻌㻌 (2)
‫݅׊‬ǡ ‫ݐ‬ൣ‫ܿ݋ݎ݌‬௧ǡ௠ǡ௜ ൌ σ௝‫א‬௃ ‫ݔ‬௧ǡ௟ǡ௜ǡ௝ ൧㻌㻌 (3)
‫݅׊‬ǡ ‫ݐ‬ൣ‫ܿ݋ݎ݌‬௧ǡ௠ǡ௜ ൑ ‫݌ܽܿ݌‬௧ǡ௠ǡ௜ ൧㻌㻌 (4)
‫ݐ׊‬ǡ ݈ǡ ݆ൣσ௜‫א‬ூ ‫ݔ‬௧ǡ௟ǡ௜ǡ௝ ൑ ‫݌ܽܿ݌‬௧ǡ௟ǡ௝ ൧㻌㻌 (5)
‫ݐ׊‬ǡ ݉ǡ ݅ൣ‫ݏ‬௧ǡ௠ǡ௜ ൑ ‫݌ܽܿݏ‬௧ǡ௠ǡ௜ ൧㻌㻌 (6)
The terms of objective function are sales of the items
recovered from item m subtracted by transportation cost,
disassembly or dismantling cost of item m, storage cost of item
m, disposal cost of item m, and purchase cost of item m,
respectively.
Constraint set (2) indicates that each provider has three
options for its purchased (or carried over) item: selling after
disassembly (or without disassembly), storing as inventory, and
discarding as waste. Constraint set (3) indicates the amount of
materials is preserved before and after each disassembly or
dismantling process. Constraint sets (4) and (5) indicate that the
total amount of the items assigned to each firm cannot exceed
its processing capacity. Constraint set (6) ensures that the
inventory of provider is smaller than its storage capacity. Any
firm corresponding to a leaf item directly reuses (or recycle) it
without any further disassembly operations. The processing
capacity of these firms is regarded as the final demand for the
item.
4. Illustration of the method
A calculation example of a mobile phone, of which RRT and
PRN are respectively given in figures 2 and 3, is explained in
this section.
4.1. Development of the linear programing model
Mobile phone recycling firms M1 and M2 are assumed to
collect 30,000 and 10,000 units of mobile phones in this period,
respectively. As a mobile phone is generally discarded as non-
burnable refuse in Japan, its purchase cost is assumed to be 0
[yen/unit]. Table 1 summarizes the storage capacity, processing
capacity, processing cost of each firm, which are tagged to each
node in the PRN in Fig. 3.
4.2. Identification of a profitable reuse and recycling plan
Among three possible disassembly or dismantling processes
in Fig. 2, two processes are estimated as profitable. The reuse
and recycling plan is (1) disassembly of a mobile phone into
LCD and PCB, (2) dismantling of PCB into rare metal fraction
and PCB’ (i.e., the PCB which is dismantled rare metal rich
component). LCDs, PCB’s, and rare metal fractions are to be
sold without any further disassembly operations and are
regarded as leaf items. The processing capacity for these items
is regarded as demand for them in this period.
 Shinsuke Kondoh et al. / Procedia CIRP 61 (2017) 245 – 250 249

4.3. Assignment of reuse/recycling processes recycling of mobile phones are generated through discussion
among multiple experts as shown in Table 3. The first one aims
Based on the reuse and recycling plan determined in section at raising the demand for second hand LCD by developing new
4.2, the optimal amount of mobile phones and PCBs that are market through provision of value-added service (e.g., technical
disassembled or dismantled in their corresponding firms are support and quality assurance). The second and the third ones
computed as well as the amount of LCDs, PCB’s, and rare metal aim at increasing the revenue from the sales of rare metal
fraction that are to be directly sold. Table 2 summarizes the fraction. The former tries to decrease the dismantling cost of
amount of LCDs and PCBs that are disassembled from mobile rare metal fraction through redesign of PCB, while the latter
phones and sold to their corresponding firms: L1 and L2, and tries to raise the price of the fraction through improving its
P1 and P2, respectively. For example, a mobile phone recycling purity by introducing new dismantling technologies.
firm M2 disassembles 5,000 units in this period and sells 5,000 Among them, the first one can make significant increase in
units of LCDs and 5,000 units of PCBs to L2 and P2, the profit. The reduction in dismantling cost of rare metal
respectively. The amount of mobile phones that are stored in fraction cannot make the process profitable in this case. Rather
this period is calculated as 5,000 units. This is because the total than that, quality improvement of rare metal fraction is more
of processing capacity of L1 and L2 (i.e., the demand for second effective in this case.
hand LCD) is smaller than the total amount of mobile phones
that are collected at M1 and M2. Insufficient demand for LCD 5. Discussion
is regarded as a bottleneck in this case.
Through the what-if analysis of the influential RRT and PRN
4.4. Calculation of objective function parameters (i.e., demand for second hand LCD, dismantling
cost of PCB, and the price of rare metal fraction), a set of ideas
The total profit of recycling and reuse of mobile phones in that is to be achieved through redesign of both products and
this period is calculated by aggregating those of mobile phones processes were identified and evaluated with the proposed
(36,860,000 [yen]) and PCBs (-43,500 [yen]). Although model. However, the generated ideas depend on implicit
dismantling of PCB are estimated as profitable in section 4.2, it knowledge of each expert. Explicit description and collection
is revealed to be non-profitable taking into account its of such knowledge as interdependency information between
associated transportation and dismantling cost. The negative RRT/PRN parameters and product and process design
balance between the sales of rare metal fraction and its parameters will be promising for further improvement of the
associated cost is also regarded as a bottleneck in this case. method. Through interpretation of such interdependency, a set
of design rules can be derived. The interdependency graph
Table. 1. Storage and processing capacity of each firm depicted in Fig. 1 can be used as a template to this end.
㻿㼠㼛㼞㼍㼓㼑㻌㼏㼍㼜㼍㼏㼕㼠㼥 㻼㼞㼏㼑㼟㼟㼕㼚㼓㻌㼏㼍㼜㼍㼏㼕㼠㼥 㻼㼞㼛㼏㼑㼟㼟㻌㼏㼛㼟㼠
㻲㼕㼞㼙 㻼㼞㼛㼏㼑㼟㼟㼕㼚㼓㻌㼕㼠㼑㼙
㼟㼏㼍㼜 㼠㻘㼙㻘㼕 㻌㼇㼡㼚㼕㼠㼉 㻌 㼜㼏㼍㼜 㼠㻘㼙㻘㼕 㻌㼇㼡㼚㼕㼠㼉 㻌 㼜㼏 㼠㻘㼙㻘㼕 㻌㼇㼈㻛㼡㼚㼕㼠㼉
The reuse and recycling plan determined in step 2 of the
㻹㻝 㻹㼛㼎㼕㼘㼑㻌㼜㼔㼛㼚㼑 㻟㻜㻜㻜 㻡㻜㻜㻜㻜㻜 㻝㻜 method is not necessarily optimal because it doesn’t consider
㻹㻞 㻹㼛㼎㼕㼘㼑㻌㼜㼔㼛㼚㼑 㻝㻜㻜㻜㻜 㻝㻜㻜㻜㻜 㻝㻜 its associated cost. In addition, there exist multiple reuse and
㻼㻝 㻼㻯㻮 㻡㻜㻜㻜㻜 㻡㻜㻜㻜㻜㻜 㻝
㻼㻞 㻼㻯㻮 㻝㻜㻜㻜㻜㻜 㻡㻜㻜㻜㻜㻜 㻝
recycling sequences corresponding to different RRTs for a
㻸㻝 㻸㻯㻰 㻜 㻝㻡㻜㻜㻜 㻜 given product in general. Thus, the identification of most
㻸㻞 㻸㻯㻰 㻜 㻞㻜㻜㻜㻜 㻜 profitable plans among all possible set of RRTs is essential to
㻼㻓㻝 㻼㻯㻮㻓 㻜 㻟㻜㻜㻜㻜 㻜
㻼㻓㻞 㻼㻯㻮㻓 㻜 㻟㻜㻜㻜㻜 㻜 improve the method and will be included in future work.
㼀㻝 㻾㼍㼞㼑㻌㼙㼑㼠㼍㼘㻌㼒㼞㼍㼏㼠㼕㼛㼚 㻜 㻟㻜㻜㻜㻜 㻜 The optimal amount of the items to be reused or recycled is
㼀㻞 㻾㼍㼞㼑㻌㼙㼑㼠㼍㼘㻌㼒㼞㼍㼏㼠㼕㼛㼚 㻜 㻟㻜㻜㻜㻜 㻜
computed from the holistic viewpoint in the method. The profit
Table 2. The amount of LCD and PCB that are processed in each firm
of some firms could be negative even when the total profit over
㻾㼑㼏㼑㼜㼠㼛㼞
㻸㻯㻰㻌㼇㼡㼚㼕㼠㼉 㻼㻯㻮㻌㼇㼡㼚㼕㼠㼉 all firms is maximized. Comparison of the optimization result
㻸㻝 㻸㻞 㻼㻝 㻼㻞 with the one maximizing profitability of each firm will be
㻹㻝 㻝㻡㻜㻜㻜 㻝㻡㻜㻜㻜 㻝㻜㻜㻜㻜 㻞㻜㻜㻜㻜
㻼㼞㼛㼢㼕㼐㼑㼞 fruitful for designing business models that promote more
㻹㻞 㻜 㻡㻜㻜㻜 㻜 㻡㻜㻜㻜
Table 3. Product and process redesign ideas effective utilization of urban mines by compensating negative
㻼㼞㼛㼒㼕㼠 profit of some firms. The extension of the method for ensuring
㻰㼕㼟㼍㼟㼟㼑㼙㼎㼘㼥㻌㼛㼒
㻳㼑㼚㼑㼞㼍㼠㼑㼐㻌㼕㼐㼑㼍㼟 㻰㼕㼟㼙㼍㼚㼠㼘㼕㼚㼓㻌㼛㼒
㼙㼛㼎㼕㼘㼑㻌㼜㼔㼛㼚㼑㼟
㻼㻯㻮㼟㻌㼇㼥㼑㼚㼉
㼀㼛㼠㼍㼘䚷㼇㼥㼑㼚㼉 the profitability of each firm will be included in future work.
㼇㼥㼑㼚㼉
㻾㼍㼕㼟㼕㼚㼓㻌㼠㼔㼑㻌㼐㼑㼙㼍㼚㼐㻌㼒㼛㼞㻌㼟㼑㼏㼛㼚㼐㻌㼔㼍㼚㼐㻌㻸㻯㻰 A linear programing model is used as an optimization tool
㻝
㼒㼞㼛㼙㻌㻟㻡㻘㻜㻜㻜㻌㼇㼡㼚㼕㼠㻛㼙㼛㼚㼠㼔㼉㻌㼠㼛㻌㻠㻜㻘㻜㻜㻜
㼇㼡㼚㼕㼠㻛㼙㼛㼚㼠㼔㼉㻌㼠㼔㼞㼛㼡㼓㼔㻌㼚㼑㼣㻌㼙㼍㼞㼗㼑㼠
㻠㻞㻘㻜㻠㻜㻘㻜㻜㻜 㻙㻡㻢㻘㻜㻜㻜 㻠㻝㻘㻥㻤㻠㻘㻜㻜㻜 for the problem in the paper. Although there exist many fast
㼐㼑㼢㼑㼘㼛㼜㼙㼑㼚㼠 algorithm to solve linear programming problem, the computing
㻾㼑㼐㼡㼏㼕㼚㼓㻌㼠㼔㼑㻌㼐㼕㼟㼙㼍㼚㼠㼘㼕㼚㼓㻌㼏㼛㼟㼠㻌㼛㼒㻌㻼㻯㻮㻌㼒㼞㼛㼙㻌㻝
㻞 㼇㼥㼑㼚㻛㼡㼚㼕㼕㼠㼉㻌㼠㼛㻌㻜㻚㻝㻌㼇㼥㼑㼚㻛㼡㼚㼕㼠㼉㻌㼠㼔㼞㼛㼡㼓㼔㻌㼞㼑㼐㼑㼟㼕㼓㼚 㻟㻢㻘㻤㻢㻜㻘㻜㻜㻜 㻙㻝㻞㻘㻜㻜㻜 㻟㻢㻘㻤㻠㻤㻘㻜㻜㻜 time increases exponentially as the number of parameter
㼛㼒㻌㻼㻯㻮㻚 increases in general. Hence, a method to adequately control the
㻾㼍㼕㼟㼕㼚㼓㻌㼠㼔㼑㻌㼜㼞㼕㼏㼑㻌㼛㼒㻌㼞㼍㼞㼑㻌㼙㼑㼠㼍㼘㻌㼒㼞㼍㼏㼠㼕㼛㼚㻌㼒㼞㼛㼙㻌㻡
㻟 㼇㼥㼑㼚㻛㼡㼚㼕㼠㼉㻌㼠㼛㻌㻝㻜㻌㼇㼥㼑㼚㻛㼡㼚㼕㼠㼉㻌㼠㼔㼞㼛㼡㼓㼔㻌㼕㼙㼜㼞㼛㼢㼕㼚㼓 㻟㻢㻘㻤㻢㻜㻘㻜㻜㻜 㻝㻟㻝㻘㻡㻜㻜 㻟㻢㻘㻥㻥㻝㻘㻡㻜㻜 problem size is necessary to apply linear programing model. In
㼕㼠㼟㻌㼜㼡㼞㼕㼠㼥
addition, linear programing model can only deal with linear
problem. Therefore, the phenomenon such as scale of economy
and learning effect, which sometimes significantly reduce
4.5. What-if analysis for generating design ideas processing cost, cannot be modeled directly by using linear
programing model. Thus, the selection of the adequate set of
For the bottlenecks identified in sections 4.3 and 4.4, three models that suit well to the context of product and
ideas that improve the economic feasibility of reuse and
250 Shinsuke Kondoh et al. / Procedia CIRP 61 (2017) 245 – 250
reuse/recycling network is quite important for interlinking
multiple decision variables over different life cycle stages.
More formal problem formulation about how to select adequate
set of optimization models is pursued in future work.
The purchase cost or sales price of each item will change
over time as well as its demand, inventory level, and
disassembly/separation cost at each firm. Thus, it is necessary
to apply the method for a series of periods with a forecast for
these parameters. The application of Life Cycle Simulation
(LCS) methods [11], which calculate every material, energy,
and monetary flow relating to a given product throughout its
life cycle, is promising to obtain such forecast. The integrated
utilization of the proposed model and LCS will be perused in
future work.
Although the paper especially focus on economic aspect to
evaluate the desirability of reuse/recycling network planning as
well as product structure, the proposed method is also
applicable to find out most environmentally friendly ones using
environmental load (e.g., energy and material consumption) as
objective functions. Simultaneous optimization from both of
environmental and economic viewpoints will be treated in
future work.
6. Conclusion
The paper proposed a linear programming model that helps
to find out optimal reuse/recycling plan of EOL products
through simultaneous consideration of product structure,
recycling processes, and reverse supply chain network.
Through a simplified case study of a mobile phone recycling,
the possibility of the model for interlinking these factors is
confirmed. Thus, the paper concludes that the proposed method
is effective for the planning of reuse/recycling of EOL products.
Future work includes following topics;
1. Acquisition of design rules for improving the
profitability of reuse/recycling of EOL products
through interpretation of analysis result of existing
cases.
2. Comparison of optimization results with those focusing
on the different aspects (e.g., material circulation rate,
carbon footprint etc.).
3. Time series calculation and optimization of the
proposed model to obtain appropriate design of product
as well as optimal reuse/recycling process planning in
the long run.
Acknowledgements
This research is partially financially supported by Grant-in-
Aid for Scientific Research (No.24603039), JSPS, Japan.
References
[1] Duflou JR, Seliger G, Kara S, Umeda Y, Ometto A, Willems B (2008)
Efficiency and Feasibility of Products Disassembly: A Case-Based Study.
CIRP Annals-Manufacturing Technology 57(2):583-600
[2] Boothroyd G, Alting L, Keynote Paper: Design for Assembly and
Disassembly. Annals of the CIRP 41(2):625-636 (1992).
[3] Knight WA, Sodhi M (2000) Design for Bulk Recycling: Analysis of
Materials Separation. Annals of the CIRP 49(1):83-86.
[4] Santochi M, Dini G, Failli F (2002) Computer Aided Disassembly
Planning: State of the Art and Perspectives. CIRP Annals -Manufacturing
Technology 51(2):507-529.
[5] Kara S, Pornprasitpol P, Kaebernick H (2006) Selective Disassembly
Sequencing:A Methodology for the End-of-Life Products. Annals of the
CIRP 55(1):37-40.
[6] Sutherland JW, Jenkins TL, Haapala KR (2010) Development of a cost
model and its application in determining optimal size of a diesel engine
remanufacturing facility. CIRP Annals-Manufacturing Technology
59(1):49-52
[7] Lee MJ, Rahimifard S (2012) A novel separation process for recycling of
post-consumer products. CIRP Annals-Manufacturing Technoloy
61(1):35-38
[8] Grunow M, Gobbi C (2009) Designing the reverse network for WEEE in
Denamark. CIRP Annals-Manufacturing Technology 58(1):391-394
[9] Sundin E. and Bras B. (2005) Making Functional Sales Environmentally
and Economically Beneficial through Product Remanufacturing. Journal of
Cleaner Production, Vol. 13, Issue 9, pp 913-925.
[10] Sundin E., Lindahl M. and Ijomah W. (2009) Product design for
product/service systems - design experiences from Swedish industry,
Journal of Manufacturing Technology Management, Vol. 20, Issue 5, pp
723-753.
[11] Takata S, Kimura T (2003) Life Cycle Simulation System for Life Cycle
Process Planning. CIRP Annals - Manufacturing Technology 52(1):37-40.