Abstract— Recently years, it is becoming important to build a

"Closed Loop Supply Chain (CLSC)" that collects used

products and reuses them. A CLSC is a way of recovering used

products from the market and reusing parts to reduce

production costs and improve the profit of the entire supply

chain. Also, the Product Life Cycle has been shortened due to

diversification of products and demand management becomes

necessary to make total sales profit.

Therefore, I thought two stage demand management with

recycle parts in CLSC for Extending product life cycle is

promising way. In fact, it is difficult to control the second

demand amount for profit and the collection rate to satisfy it, so

this research on addressing these problems is effective. Further,

I formulate the method for the production planning with two

stage demand management by Mixed Integer Programing

optimization model. And maximize SCM total profit by

controlling recovery rate and second demand start time.

Index Terms— Closed Loop Supply Chain (CLSC), Demand

Management, Product Life Cycle, Reverse Logistics.

I. INTRODUCTION

N recent years, the demand for resources has increased.

Also "Environmental protection" and “Resource saving”

are becoming important. Therefore, there is a need to

establish a Closed Loop Supply Chain (CLSC) that expands

profits by reusing recycled parts obtained from it (Fig1).

Also, the Product Life Cycle has been shortened due to

diversification of products and demand management

becomes necessary to make total sales profit. Therefore, I

thought two stage demand management with recycle parts in

CLSC for Extending product life cycle is promising way

(Fig2). In fact, there are few studies on CLSC, there is also no

research combined with demand management in previous

research. In addition, it is difficult to control the second

demand amount for profit and the collection rate to satisfy it,

so this research on addressing these problems is effective. ManufacturerNew Parts Supplier

Market

Reused Parts Supplier

Demand

Disposal

Collection

Order

Order

Received order

Received order

Fig1. Closed Loop Supply Chain

Ryuhei Kajiyama, Graduate School of Information, Production and

Systems Waseda University, Japan (e-mail : ryuhei0554@akane.waseda.jp)

Tomohiro Murata, Professor Graduate School of Information, Production

and Systems Waseda University, Japan (e-mail : t-murata@waseda.jp)

0

100000

200000

300000

400000

500000

600000

700000

800000

2018/1

/1

2018/2

/1

2018/3

/1

2018/4

/1

2018/5

/1

2018/6

/1

2018/7

/1

2018/8

/1

2018/9

/1

2018/1

0/1

2018/1

1/1

2018/1

2/1

2019/1

/1

2019/2

/1

2019/3

/1

2019/4

/1

2019/5

/1

2019/6

/1

2019/7

/1

2019/8

/1

2019/9

/1

2019/1

0/1

2019/1

1/1

2019/1

2/1

2020/1

/1

2020/2

/1

2020/3

/1

2020/4

/1

2020/5

/1

2020/6

/1

2020/7

/1

2020/8

/1

2020/9

/1

2020/1

0/1

2020/1

1/1

2020/1

2/1

Demand1 amount(Normal) Demand2 amount(Lower)

Demand

Management

Extending PLC

Stage1

Stage2

(Demand)

(Month) Fig2. Image of two stage demand management

II. RESEARCH GOAL

In this research, proposing a novel production planning

method for two stage demand management with recycle parts

in CLSC model to maximize total sales profit.

In addition, considering the “collection rate” of the

recovery product from the initial demand and the “sale start

time” of the second demand in the CLSC model, the objective

is to find a combination for max profit of the entire supply

chain due to each fluctuation.

Also, formulating the model for the production planning

with two stage demand management by Integer Programing

optimization model and evaluating the proposed model with

case study to find feasible condition to increase max profit.

Through four experiments, I examined the effectiveness of

the model proposed in this research.

III. SURVEY OF RELATED RESEARCH AND DRAWBACKS

Summarized the previous research on the Closed Loop

Supply Chain as follows. Kaya studied incentives and

product reproduction. Samar studied the reproduction model

considering the quality of recycled parts. Watanabe [1]

collected used products by paying incentives and decided the

optimal order amount according to the quality level of the

collected products. Kainuma [2] is doing a lot of research on

the recycling type supply chain, and He studied the

effectiveness of VMI under the closed supply chain. Kaya [3]

studied incentives and product reproduction. Samar [4]

studied the reproduction model considering the quality of

recycled parts.

However, there were no studies reusing parts recovered by

CLSC for two stage demand.

PROBLEM DESCRIPTION

A. Closed Loop Supply Chain model

The CLSC model proposed in this research is as shown

in the figure (Fig3). The CLSC model consists of one

Production Planning for Closed Loop Supply

Chain with Recycle Parts and Product Life Cycle

Ryuhei Kajiyama and Tomohiro Murata.

I

Proceedings of the International MultiConference of Engineers and Computer Scientists 2019 IMECS 2019, March 13-15, 2019, Hong Kong

ISBN: 978-988-14048-5-5 ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

IMECS 2019

manufacturer, one new parts supplier, and one recycled

parts supplier. Manufacturers produce two products,

“Normal Priced Product” and “Lower Price Product”.

First of all, “Normal Priced Product” is distributed to the

market. Later, these products are recycled by recycled parts

supplier from the market by collection rate “c”. Recycled

parts supplier disassembles and inspect the collected

products. Recycled parts are reused as “Lower Priced

Product”. Also, when demand can’t be satisfied by recycled

parts, it is procured new parts and manufacture products.

demand

Collection of Recycled Products

New

PartsDisposal

ManufacturerNew Parts Supplier

Recycled Parts Supplier Demand2

Normal priced

products

Lower priced

products

Demand1

Disposal

Inspection

Optimal

Ordering plan

customer

Recycled

PartsRecycle

(Rate : c)

Fig3. Closed Loop Supply Chain model

B. Initial Demand Model (Bass Model)

Initial demand is expressed by using “Bass model”. The

result is as shown in the figure below (Fig4).

The Bass model is a demand forecasting model formulated

by adding customer's purchasing behavior to the logistic

curve. It is a model suitable for simulating the diffusion

process of durable consumer products in particular. The Bass

model is formulated as follows and is composed of

“Maximum cumulative sales amount: m”, “External

influence factor: p”, “Internal influence factor: q”, and

“Time: t”.

2)()(2 )/()(1 tqptqp

t qepeqpmD (1)

C. Collection Amount Model (Bass Model)

The amount of recycle products from the initial demand is

determined by the following formulation. It is formulated by

adding “Collection Start Time: Δ t” and “Collection Rate: C”

to the initial demand model. (Fig4)

)(1

ttt DcCL ⊿ (2)

0

100000

200000

300000

400000

500000

600000

700000

800000

Demand1 amount(Normal)

Recycle available amount

Δ t

(Demand)

(Time)

Fig4. Initial Demand and Collection Amount Curve

D. Second Demand Model (Bass Model)

The second demand is determined from the formulation

that when Sales Start Time “s” is delayed, the total demand

amount also decreases by “d” (Fig5). There is the following

trade-off by adding conditions.

When sales start time is early, demand is high, but

inventory of recycled parts is small and manufacturing price

increases. When sales start time is later, demand decrease and

it can’t be got profit.

))(1(})/()({2 2)()(2

, stdqepeqpmD s

tqptqp

st δ

0

50000

100000

150000

200000

250000

300000

20

18

/1/1

20

18

/2/1

20

18

/3/1

20

18

/4/1

20

18

/5/1

20

18

/6/1

20

18

/7/1

20

18

/8/1

20

18

/9/1

2018/1

0/1

2018/1

1/1

2018/1

2/1

20

19

/1/1

20

19

/2/1

20

19

/3/1

20

19

/4/1

20

19

/5/1

20

19

/6/1

20

19

/7/1

20

19

/8/1

20

19

/9/1

20

19

/10

/1

20

19

/11

/1

20

19

/12

/1

20

20

/1/1

20

20

/2/1

20

20

/3/1

20

20

/4/1

20

20

/5/1

20

20

/6/1

20

20

/7/1

20

20

/8/1

20

20

/9/1

2020/1

0/1

2020/1

1/1

2020/1

2/1

(Demand)

(Time) Fig5. Second Demand Curve

E. Max Profit Function

The formulation for obtaining the maximum profit is as

follows.

Max Profit = Total sales of Lower priced Products

- (Procurement cost of Recycle parts + Procurement cost of

New parts) - Inventory cost of Recycle parts

IV. FORMULATION OF THE PROBLEM

A. Symbol

I set symbols for max profit function (Table1).

Object Function for Max Profit

I formulated a function that decides “Optimal Sales Start

time” for maximize profit. The expression is as follows. Also,

calculate as below the procurement amount of recycled parts

and inventory amount.

TABLE I

SYMBOL FOR MAX PROFIT FUNCTION

Symbol Quantity

D1 Demand amount of “Normal Priced Products” at time t

D2 m

p

q

d

Demand amount of “Lower Priced Products” at time t

Total demand amount

Prospect of starting to use it under the influence of Ad

It is expected that the products will be affected by people

Demand reduction rate

ILt Inventory amount of “Lower Priced Products” at time t

ILC Inventory cost of “Lower Priced Products”

CRt Amount of available recycle parts at time t

c Collection rate of recycle products

PNt Procurement amount of new parts

PRt Procurement amount of recycle parts

PNC Procurement cost of new parts

PRC Procurement cost of recycle parts

PL Price of “Lower Priced Products”

δs Decision making [0, 1] of sale start time

Proceedings of the International MultiConference of Engineers and Computer Scientists 2019 IMECS 2019, March 13-15, 2019, Hong Kong

ISBN: 978-988-14048-5-5 ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

IMECS 2019

36

1

36

1

36

1

, )(2

)(

t t

ttt

t

st PRPRCPNPNCILILCDSPL

sPROTFIT

ttttt ILDDDPR }3/)222({ )1()1(

)2()1( ttt PRILIL

V. METHOD TO SOLVE THE PROBLEM

This research is solved by the following method. While

changing the values of “collection rate” and “sales start

time” to the formulated CLSC model, find the combination

that will maximize profit under the given parameters. Also,

simulation is done by coding in R language.

Formulation of Optimization Model

Set Parameter

Collection Rate Optimal sales start time

Analysis of optimal solution for profit

maximization by “GNU R”

Sensitivity Analysis

I decide the “Optimal sales start time” and “Optimal procurement

amount” for maximizes profits and find the most influential factors.

Fig6. Method to solve of my research

VI. EXPERIMENT

In this research, conducting “Four experiments”. The

contents of the experiment are as follows. In addition, each

parameter was set as shown in Table 2. Similar values were

used in all experiments.

A. Experiment1

Finding “Optimal sales start time” and “Procurement plan

and Inventory plan of recycle parts” under the given

conditions. In this experiment, I set the collection rate as

“0.5”. The result is as follows (Fig7).

(8,000,000,000)

(6,000,000,000)

(4,000,000,000)

(2,000,000,000)

0

2,000,000,000

4,000,000,000

6,000,000,000

8,000,000,000

10,000,000,000

12,000,000,000

14,000,000,000

-510000

-310000

-110000

90000

290000

490000

690000

890000

20

18/1

/1

20

18/2

/1

20

18/3

/1

20

18/4

/1

20

18/5

/1

20

18/6

/1

20

18/7

/1

20

18/8

/1

20

18/9

/1

20

18/1

0/1

20

18/1

1/1

20

18/1

2/1

20

19/1

/1

20

19/2

/1

20

19/3

/1

20

19/4

/1

20

19/5

/1

20

19/6

/1

20

19/7

/1

20

19/8

/1

20

19/9

/1

20

19/1

0/1

20

19/1

1/1

20

19/1

2/1

20

20/1

/1

20

20/2

/1

20

20/3

/1

20

20/4

/1

20

20/5

/1

20

20/6

/1

20

20/7

/1

20

20/8

/1

20

20/9

/1

20

20/1

0/1

20

20/1

1/1

20

20/1

2/1

Profit Demand1 amount(Normal) Recycle available amount Demand2 amount(Lower)

(Demand)

(Month)

(Profit)

Demand1

Demand2

Collection Amount

Profit

Fig7. Results under the given condition

Firstly, the reason why the profit became negative was that

the recycled parts were not recovered enough to satisfy the

demand, and because the procurement costs took to satisfy

the demand with new parts, it can be said that this resulted.

0

50000

100000

150000

200000

250000

300000

350000

400000

201

8/1

/1

201

8/2

/1

201

8/3

/1

201

8/4

/1

201

8/5

/1

201

8/6

/1

201

8/7

/1

201

8/8

/1

201

8/9

/1

2018/1

0/1

2018/1

1/1

2018/1

2/1

201

9/1

/1

20

19

/2/1

20

19

/3/1

20

19

/4/1

20

19

/5/1

20

19

/6/1

20

19

/7/1

20

19

/8/1

20

19

/9/1

20

19

/10

/1

20

19

/11

/1

20

19

/12

/1

20

20

/1/1

20

20

/2/1

20

20

/3/1

20

20

/4/1

20

20

/5/1

20

20

/6/1

20

20

/7/1

20

20

/8/1

20

20

/9/1

20

20

/10

/1

20

20

/11

/1

20

20

/12

/1

Demand2 amount(Lower) Invwntory amount ProcurementL

(Demand)

(Month)

Procurement

Inventory

Demand2

Fig8. Behavior of Inventory and Procurement Recycle Parts

The above figure shows the behavior of inventory amount

and recovered amount of recycled parts.

When inventory amount increased more than demand, it was

such a result because recycled parts were conditioned not to

procure.

B. Experiment2

In this experiment, finding the combination of the sales

start time “s” and collection rate “c” for max profit. In

experiment 1, I decided the optimal sales start time under

given conditions. In experiment 2, changing the sales start

time and the collection rate of recycled parts and finds the

combination with the maximum profit (Fig9).

0.1

0.3

0.5

0.7

0.9

0

50000000000

100000000000

150000000000

200000000000

250000000000

7m

on

th

8m

on

th

9m

on

th

10

mo

nth

11

mo

nth

12

mo

nth

13

mo

nth

14

mo

nth

15

mo

nth

16

mo

nth

17

mo

nth

18

mo

nth

start time 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

(Month)

(Collection rate)

(Profit)

Max Profit

Fig9. Combination of the sale start time and collection rate

Overall, when collection rate was small and sale start time

was delayed, max profit decreased. The reason is that when

the total demand amount decreased, the sales start time was

delayed and max profit decreased also.

Similarly, the lower the recovery rate, the lower the overall

profit. It is thought that this is because the collection amount

can’t satisfy the demand amount and it was supplemented

with new parts, so it took procurement cost.

TABLE 2

SET PARAMETERS

Symbol Quantity Value

T Total Life Cycle 36 (Month)

p Innovation effect for Demand1 0.03

q Imitation effect for Demand1 0.5

m

p

q

m

Total amount of Demand1

Innovation effect for Demand1

Imitation effect for Demand1

Total amount of Demand1

5000000

0.06

0.25

5000000

c Collection Rate 0.1 ~ 0.9

PNC Procurement Cost of New Parts 70000

PRC

Procurement Cost of Recycle Parts 40000

ILC Inventory Cost of recycle Parts 10000

SPL Sale Price of Lower Priced Product 100000

.

Proceedings of the International MultiConference of Engineers and Computer Scientists 2019 IMECS 2019, March 13-15, 2019, Hong Kong

ISBN: 978-988-14048-5-5 ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

IMECS 2019

160000000000

165000000000

170000000000

175000000000

180000000000

185000000000

190000000000

195000000000

200000000000

205000000000

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1

8month

(Collection Rate)

(Profit)

Fig10. Relationship Max Profit and Collection Rate

when collection rate is high, inventory cost increased, and

when collection rate is small, because procurement costs of

new parts are required, it is considered that the maximum

profit was obtained at the collection rate of 0.5. At the time of

this combination the maximum profit could be obtained

“54322876667”, which was about 15.7% better than when no

second demand management was done (Fig11).

0

50000000000

100000000000

150000000000

200000000000

250000000000

300000000000

350000000000

400000000000

450000000000

Only initial demand two stage demand management

(Max Profit)

15.7%

Fig11. Profit improvement rate

C. Experiment3

In this experiment, finding second demand boundary for

getting profit by two stage demand management. In

Experiments 1 and 2, the second demand was set to be less

than 50% of the initial demand. By doing this experiment, I

find the required demand for profit. By finding this necessary

demand amount, it becomes a criterion as to whether or not

profit will be obtained when forecasting demand. The

conditions of the experiment are the combination of the sales

start time and the collection rate for the maximum profit.

(20,000,000,000)

0

20,000,000,000

40,000,000,000

60,000,000,000

80,000,000,000

100,000,000,000

500

000

0

400

000

0

300

000

0

200

000

0

190

000

0

180

000

0

160

000

0

156

000

0

155

500

0

155

480

0

155

475

0

155

473

0

155

472

7

155

472

0

155

470

0

155

460

0

155

450

0

155

400

0

155

000

0

150

000

0

100

000

0

Profit

(Demand)

(Max Profit)

Boundary of Demand Amount

C=0.5

Fig12. Boundary of “Max Profit” and “Demand Amount”

Above figure shows the results of this experiment. The

experiment started with the same amount of demand as the

initial demand, and gradually decreased from there. As a

result, I found that the required demand amount will be

profitable even if it is even smaller than the demand amount

set initially.

D. Experiment4

In this experiment, finding recycled parts cost boundary to

get total profits by two stage demand management.

Experiment 4 was performed under the same conditions as in

Experiment 3. By finding the boundary of recycle parts cost,

it is possible to increase profits. The cost of recycled parts

was experimented from the same price as the new parts, and

the cost was gradually reduced. When the price was

equivalent to that of the new part, it was impossible to obtain

profits as expected from the experiment. The price at which

profit actually appears is about 3 times the price set at the

beginning. The following figure is the result of this

experiment, the circled position is the boundary.

(12,000,000,000)

(10,000,000,000)

(8,000,000,000)

(6,000,000,000)

(4,000,000,000)

(2,000,000,000)

0

2,000,000,000

Max Profit

(Recycle Parts Cost)

(Max Profit)

Boundary of Recycle Parts Cost

C=0.5

Fig13. Boundary of “Max Profit” and “Recycle Parts Cost”

VII. CONCLUSION

In this research, I created “Two Stage Demand” by

creating CLSC model and reusing recycled parts, and

expanded the profit. In addition, I formulated the CLSC

model and simulated under constraint conditions and

conducted five experiments. By coding in R language and

changing parameters, I was able to satisfy all experiments.

Through four experiments I found a combination for the

maximum profit of “collection rate” and “sales start time”

“lead time of recycle parts”. I also found boundaries of

recycled parts prices and demand amount for maximum

profit.

However, I’ve already known that how much “demand

amount” by Bass Model in this research, so I can say that it

was a slightly different environment from the real problem.

For that reason, I would like to conduct a similar

experiment under the circumstance where the amount of

demand is uncertain in the future.

REFERENCE

[1] Takeshi Watanabe (2013), “Optimal Operation for GSCM considering

collection incentive and quality for recycling of used products”, 日本

経営工学会論文誌, p226-p235.

[2] Kainuma (2009), “クローズドループサプライチェーンにおける

ハイブリッド生産/再生産システム”, p1-p4.

[3] Onur Kaya (2010), “Incentive and production decisions for

remanufacturing operations”, European journal of operational research

201, p442-p453.

[4] Samar K (2009), “Joint procurement and production decisions in

remanufacturing under quality and demand uncertainty”, production

economics 120, p05-p17.

[5] Jean-Pierre(2012), “Production planning of a hybrid manufacturing /

remanufacturing system under uncertainty within a closed-loop supply

chain”, International Journal of Production Economics, Vol.135, No.1.

[6] Yuki Ooshita (2009), “クローズド・ループサプライチェーンにお

ける生産・再生産ハイブリッドシステム”, p01-p04

Proceedings of the International MultiConference of Engineers and Computer Scientists 2019 IMECS 2019, March 13-15, 2019, Hong Kong

ISBN: 978-988-14048-5-5 ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)

IMECS 2019