ArticlePDF Available

Environmental Science and Pollution Research E-WASTE MANAGEMENT USING GAME THEORY CONCEPT Optimal recycle price game theory model for second-hand mobile phone recycling

Authors:

Abstract

Human societies develop rapidly through the advancement of technology; however, with these advancements, many problems are emerging. The topic chosen for this study surrounds the e-waste, which has become a major problem around the world. Secondhand and unused mobile phones are a big part of globally generated e-waste. If these devices are properly recycled, they can generate substantial economic and resource value. Yet if they are indiscriminately discarded, they cause a profound environmental impact. Given the current low recovery rate of mobile phones, an increase in recovery rates becomes critical in lessening economic and environmental impacts. Based on the status quo of secondhand mobile phone recycling processes in China, this article analyzes the behavior of individuals and recyclers through a comprehensive static information game theory and finds ways to increase the recycling rate of secondhand mobile phones. The study helps the customers, to clearly identify the recycle price. In case of market, the government policy can be introduced with a reward and punishment mechanism. Furthermore, under the ideological guidance of game theory, this paper also establishes a corresponding price model of secondhand mobile phone recycling based on best response dynamics like search, variable neighborhood search, and hybrid meta-heuristic method. This model shows that the recovery time differences have a significant impact on the recovery price. Moreover, to an extent, this model can promote the possibility and initiative of customers choosing cell phone recycling.
Vol.:(0123456789)
1 3
Environmental Science and Pollution Research
https://doi.org/10.1007/s11356-021-17061-w
E-WASTE MANAGEMENT USING GAME THEORY CONCEPT
Optimal recycle price game theory model forsecond-hand mobile
phone recycling
KennedyE.Ehimwenma1· SujathaKrishnamoorthy1· ZixuanLiu1· YangQiu1· YihangLiu1· WangyingDou1
Received: 1 December 2020 / Accepted: 11 October 2021
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2021
Abstract
Human societies develop rapidly through the advancement of technology; however, with these advancements, many problems
are emerging. The topic chosen for this study surrounds the e-waste, which has become a major problem around the world.
Second-hand and unused mobile phones are a big part of globally generated e-waste. If these devices are properly recycled,
they can generate substantial economic and resource value. Yet if they are indiscriminately discarded, they cause a profound
environmental impact. Given the current low recovery rate of mobile phones, an increase in recovery rates becomes critical
in lessening economic and environmental impacts. Based on the status quo of second-hand mobile phone recycling processes
in China, this article analyzes the behavior of individuals and recyclers through a comprehensive static information game
theory and finds ways to increase the recycling rate of second-hand mobile phones. The study helps the customers, to clearly
identify the recycle price. In case of market, the government policy can be introduced with a reward and punishment mecha-
nism. Furthermore, under the ideological guidance of game theory, this paper also establishes a corresponding price model of
second-hand mobile phone recycling based on best response dynamics like search, variable neighborhood search, and hybrid
meta-heuristic method. This model shows that the recovery time differences have a significant impact on the recovery price.
Moreover, to an extent, this model can promote the possibility and initiative of customers choosing cell phone recycling.
Keywords Second-hand mobile phone· Recycling· Noncooperation and cooperation· Two-player game· Static games of
complete information
Introduction
Current situation aboutelectronic waste
One of the critical global issues in the modern era is the
growing quantity of e-waste and its toxic effect on the envi-
ronment and people (Hidasan etal. (2014); Arı and Yılmaz
(2016a)). In 2019, a total of 53.6 Mt of e-waste was gener-
ated globally, calculated to 7.3 kg per capita. This figure is
expected to increase to around 74.7 Mt by 2030 (Forti etal.
(2020)). Mobile phones are large part of e-waste products.
According to Deng etal. (2017), due to mobile phones’ short
service life, which is usually less than 2~3 years, the pro-
duction of waste mobile phones is considerable. Although
consumers have a general awareness of the importance of
recycling electrical and electronic equipment including
mobile phones, their awareness has not been fully trans-
formed into recycling behavior (Ylä-Mella etal. (2015)).
Compared with traditional household waste, mobile phones
contain highly toxic substances and valuable materials, such
Responsible Editor: Philippe Garrigues
* Sujatha Krishnamoorthy
sujatha.ssps@gmail.com
Kennedy E. Ehimwenma
kehimwen@kean.edu
Zixuan Liu
zixuanl@kean.edu
Yang Qiu
yangq@kean.edu
Yihang Liu
yihangl@kean.edu
Wangying Dou
wanyingd@kean.edu
1 Department ofComputer Science, Wenzhou-Kean
University, Wenzhou, China
Environmental Science and Pollution Research
1 3
as copper, silver, gold, and palladium all which can be recy-
cled. Therefore, recycling mobile phones has a dual value of
environmental protection and resource conservation.
From the perspective of saving resources and protecting
the environment, solving the problem of how to increase the
recycling rate of mobile phones is essential.
Moreover, as the world’s largest exporter of electrical and
electronic equipment and importer of used electrical and
electronic equipment, China plays a crucial role in the treat-
ment of electronic products (Chi etal. (2011)); thus, this
article attempts to analyze how to improve the recycling of
second-hand phones in the nation..
Hicks etal. (2005) pointed out that in China and other
developing and industrial countries, waste is viewed as a
resource and income-generating opportunity. Therefore, the
market hopes that consumers and recyclers can actively par-
ticipate in recycling used mobile devices. However, how to
effectively manage e-waste in China is imminent. E-waste
management is a multi-stakeholder problem. To solve the
problem effectively, this article applies game theory in used
mobile phone recycling.
Analyzing thestructure ofclosed‑loop supply
chains: agame theory perspective
Game theory is a field of applied mathematics used to
analyze complex interactions between entities (Hadzic
etal. (2013)). Game theory’s focus is on multi-intelligent
rational decision-making regarding conflict, competition,
and cooperation. Moreover, it is a useful tool for obtaining
their corresponding benefits from each other by analyzing
each factors’ behavior and the information known (Myer-
son (2013)). Cooperative games and non-cooperative games
are two primary situations involved in game theory. These
include theories of interpersonal interaction, also known as
people and economic behavior. Cooperative game theory
studies how people distribute the benefits of cooperation
through cooperative strategy and negotiation among the
various agents and various income distribution problems.
Cooperative game theory promotes both parts’ interests
in the game and the whole of society’s interests through
cooperation. Thus, cooperative game theory emphasizes
collectivism and collective rationality. On the other hand,
non-cooperative game theory emphasizes individualism and
individual rationality; it is the study of how people choose to
maximize their interests insituations where interests affect
each other, that is, the problem of strategic choice. At pre-
sent, the two sides, consumer and company, are in a state
of non-cooperation, which is also the root of the serious
e-waste problem in China. Therefore, how to change this
situation becomes very important.
For the second-hand mobile phone recycling, a good
strategy should not only promote the interests of both
consumer and company, but also advance the interests of the
whole society. Therefore, we hope to turn the previous non-
cooperation between the two parties to cooperation through
the game theory and build an optimal recycle price model to
increase the number of second-hand mobile phone recycling
and raise consumers’ awareness of recycling.
This paper is organized with a background study in the
“Background” section. The “Application of game theory for
recycling” section mainly focuses on analysis and the litera-
ture work about the meta-heuristic algorithms and game the-
ory which plays an important role in the recycling models. In
“The ORPGT model of second-hand mobile phone” section,
our proposed method for finding the optimal recycle price
(ORP) uses game theory, and the experimental results and
discussion is elaborated upon in the “Experimental results
and discussion” section. The paper is concluded with our
findings in the final “Conclusion” section.
Background
Consumer behavior insecond‑hand mobile phone
recycling
The current research on mobile phone recycling mainly
focuses on analyzing consumer behavior and improving the
recycling management model. For consumer behavior, the
low recycling price and residents’ lack of environmental
awareness have caused failing recycling (Su (2018); Du etal.
(2014)). However, Ylä-Mella etal. (2015) state that consum-
ers are aware of waste recycling systems’ importance but
that the awareness has not been transformed into recycling
behavior. In contrast, Tan etal. (2018) further confirmed the
recycling awareness is not the main reason for recycling. It
is because financial benefits from recycling for individual
customer are too low compared to the original price. Arı
and Yılmaz (2016b) found that guidance and motivation can
effectively improve housewives’ willingness to recycle and
increase the recovery rate of mobile phones. Zhang etal.
(2019) also ascertained that consumers’ personality traits
essentially affect phone recycling. Consumer behavior influ-
ences second-hand mobile phones recycling, and positive
behavior helps to increase the recovery rate.
Recycling model based ongame theory
Different researchers have adopted various game theory
approaches to improve the recycling management model.
Tan etal. (2018) construct a complete information static
game to transform recycling pricing into the problem of
solving the Nash Equilibrium point in recycling. Liu and
Su (2019) used the game method to calculate the govern-
ment’s optimal subsidy pricing and reported that the fund
Environmental Science and Pollution Research
1 3
pricing of some kinds of abandoned electrical and electronic
products was not reasonable through data simulation. Wang
etal. (2020) established a dynamic Stackelberg game mode
and found that the reward and punishment mechanism can
significantly improve waste products’ recovery rate. Further-
more, Shi etal. (2020) proved theoretically that cooperative
games could bring more profit to consumers’ supply chain
and benefits. Nash Equilibrium between cooperative con-
sumers and enterprises is the key to improving the recov-
ery rate of used mobile phones. However, due to the inap-
propriate pricing of second-hand mobile phones, the game
between consumers and enterprises fails to be Nash equilib-
rium. Previous studies mostly started from theoretical game
theory to explore recycling and the government’s reward
and punishment mechanism to help consumers and enter-
prises approach Nash equilibrium. This study points out the
necessity of second-hand mobile phone pricing balance for
recycling through a detailed analysis of consumer behavior,
corporate behavior based on game theory, and the govern-
ment’s reward and punishment mechanism. Then it analyzes
the decisive factors of second-hand mobile phone pricing
based on machine learning. Finally, optimization methods
through Tabu Search, Variable Neighborhood search, and
hybrid meta-heuristic algorithm are established to achieve
Nash Equilibrium recycling price. Recently Shekarian and
Flapper (2021) has referred a closed-loop supply chain
(CLSC) model that is seen as one of the circular economy’s
leading approaches for reducing our natural environment
load. We will be utilizing this in future in our research.
Application ofgame theory forrecycling
Hawk‑Dove game betweenrecycler andconsumer
The following parameters are used to analyze the recycling
game between consumer, enterprise, and government in
Table1.
For companies, storage, transportation, and labor in the
recycling process will incur recycling costs, which are usu-
ally challenging to reduce. Consumers are unwilling to share
the cost of the recycling process, only 47.9% of consumers
agreed to pay for only 0~ 5% of the used mobile phone
cost (Yin etal. (2014)). Some companies decline second-
hand mobile phones’ recycling prices to reduce cost and
maximize profit. Furthermore, in the pricing of second-hand
mobile phones, companies are usually in a dominant posi-
tion. In the recycling game between individuals and enter-
prises, recyclers are reluctant to raise recycling prices and
regularly hold a non-cooperative attitude.
According to the following Hawk-Dove game (SMITH
and PRICE (1973)) between recyclers and consumers, the
recyclers have governing control over recycling prices. In the
Hawk-Dove model parameters, suppose consumers expect
the ideal recovery price to be Pi, recycling companies give
their best offer to be Po. The revenue from recycling phones
is R. Generally, Po < R < Pi.
If one is a Dove (compromise) and the other is a Hawk
(Hardliners), the transaction price is the Hawk offered, so
that the price will be Po.
If both parties are Doves, both make concessions, and
then the transaction price is Pm, Po < Pm < Pi.
If both parties are Hawks, the transaction cannot be real-
ized, the consumer has no profit, and the companies cannot
profit from recycling. Assume the revenue from recycling
phones is R, and then the model is shown in Table2.
Given the Dove and Hawk model presented in the forego-
ing section, the detailed analysis of the work is presented in
the following section.
If the consumer chooses the Dove strategy, the enterprise
insists that the hawk strategy is the optimal scheme (since
R−P> R−Pm). If consumers insist on hawk strategy, enter-
prises insist on hawk strategy is the optimal scheme (since
R − Pi < 0).
Therefore, the best choice for enterprises is to adhere to the
Hawk strategy, and when they did, consumers can only choose
the Dove strategy. Hence, based on the Hawk-Dove game, the
recovery price decision lies with the recycler, which leads to
companies’ negative attitudes toward cooperation. Besides,
even if some companies are willing to benefit consumers,
Table 1 The set of parameter
Parameter Description
RRevenue from recycling phone
PmTIf both parties are Doves the transaction price
PiConsumer ideal recycle price
P0Best recycling price that enterprise offers
PhHigher payoff for enterprises when the status
quo of used mobile phones improves
PLLower payoff for enterprises when the status
quo of used mobile phones is not improved
QThe payoff of the passive participant
qThe payoff of the active participant
DRewards for active participant
EPenalties for passive participant
Table 2 The Hawk-Dove game between recycler and consumer
Consumer
Dove Hawk
Recycler Dove (R−Pm, R−Pm) (R−Pi, Pi)
Hawk (R−P0, P0) (0, 0)
Environmental Science and Pollution Research
1 3
they will eventually be affected by other recyclers in market
competition.
Prisoner’s dilemma gaming inrecycling mobile
phones
When few recycles enterprises actively participate in recy-
cling, the negative attitude of other recyclers will weaken the
recycling. Individual rationality leads to collective irrationali-
ties, like the prisoner’s dilemma in recycling.
Recycling has both active participation and passive partici-
pation strategies. When the status quo of used mobile phones
is significantly improved, the enterprises involved in recycling
of used mobile phones will get the same higher payoff: Ph.
When the status quo of used mobile phones is not improved,
the recyclers of used mobile phones will get the same lower
payoff: PL. When one party A and B participate actively and
the other party passively participates, the payoff of the active
participant is q, and the payoff of the passive participant is Q.
The Table3 gives a How the participants are classified in the
Prisoners dilemma and the notations used.
The mobile phone recycling prisoner’s dilemma model
has two equilibrium: (Ph, Ph), (PL, PL). The (Ph, Ph ) Pareto
efficiency is better than (PL, PL). However, due to the cost of
recycling mobile phones, active participation cost is higher
than passive participation. The passive party will share the
active participant’s results without the cost, so Q > Ph. A
single recycler’s active cooperation cannot accomplish the
increase in second-hand mobile phones’ recycling rate. It
requires the association of all mobile phone recyclers. When
only a few recyclers actively participate in recycling, the
recycling rate will not be improved, and the result of the
game is the same as the passive participation of both par-
ties. Because passively participating recyclers bring negative
externalities to actively cooperating recyclers, the actively
participating party pays higher recycling costs, so q < PL.
From the perspective of personal rationality, A and B will
both choose to participate passively.
The government reward andpunishment
mechanism
To turn non-cooperation into cooperation, it is decided
to introduce the government to regulate the behavior of
enterprises. Therefore, we divide our participants into
three main parts: the government, the enterprise, and the
consumer. And the key players in the regulatory recovery
mechanism are shown in Fig.1.
Government reward and punishment mechanism refers
to the process of administrative management, the govern-
ment supervises and inspects the assigned tasks and rewards
active participants and punishes the less active in the promo-
tion of policy implementation.
Punishment mechanism: When mobile phone recyclers
participate passively, the government enforces economic
penalties -E. Its standard penalties are irrespective of the
mobile phone models or brands.
Reward mechanism: When mobile phone recyclers par-
ticipate passively, the government issues corresponding
rewards D. Table4 indicates the penalty.
After the reward and punishment mechanism is intro-
duced, active participation is a dominant strategy for recy-
clers, so the Nash Equilibrium becomes (active participation,
active participation)
The ORPGT model ofsecond‑hand mobile
phone
Human interaction is studied by using some of the bio-
inspired algorithms and game theory strategies involving
two or more persons. Hence, the paper proposes a method
to utilize Nash Equilibrium which is most widely used in
the gaming literature. The following figure gives the over-
all structure of our proposed model for finding the optimal
recycle price using the game theory Fig.2
This part of the paper explains the model that has been
used for predicting the price for recycling phones and the
parameters that been assumed for implementation. Machine
Table 3 Prisoner’s dilemma in recycling used mobile phones.
Company A/Company B Active participation Passive
participa-
tion
Active participation (Ph, Ph) (q, Q)
Passive participation (Q, q) (PL, PL)
Fig. 1 Three key players in game with and their relationship
Environmental Science and Pollution Research
1 3
learning algorithm is used to calculate the utility function
for specific phones and conditions.
Model forpredicting therecycling price
This research data contains 2275 effective instances of sec-
ond-hand mobile phones (1022 Huawei and 1253 Apple,
42 Huawei Nova, 28 Huawei Nova 2, 42 Huawei Nova 2
PLUS, 63 Huawei Nova 2s, 63 Huawei Nova 3i, 42 Huawei
Nova 3e, 21 Huawei Nova 5, 42 Huawei Nova 5 PRO, 42
Huawei Nova 5i, 84 Huawei Nova 5i pro, 63 Huawei Nova
6, 21 Huawei Nova 6 SE, 84 Huawei P20 Pro, 42 Huawei
Mate 20, 42 Huawei Mate 20 Pro, 63 Huawei Mate 20 X,
70 Huawei Mate 30, 105 Huawei Mate 30 Pro, 63 iPhone
7, 42 iPhone 8, 63 iPhone 7 Plus, 252 iPhone 11 Pro, 252
iPhone 11 Pro Max, 189 iPhone 11, 63 iPhone XS, 63
iPhone XS Max, 126 iPhone XR, 63 iPhone 8 Plus, 63
Huawei P30 Pro, and 77 iPhone SE2) and covers 13 attrib-
utes like mobile communication technologies (includes 4G
and 5G phone), brand name of the company, phone name,
version, RAM size, ROM, time difference (length of time
between purchase and recycling), CPU, operating system,
screen size (inches), screen resolution, and level. The level
is fixed with some criteria such as the appearance and
physical damages caused during the recycling process.
The following are the 6 levels used in game theory for
recycling purposes.
Level 1: The appearance is perfect, and the screen is free of
scratches.
Table 4 Penalty matrix Without the reward and punishment mechanism
Company A/Company B Active participation Passive participation
Active participation D, D D, -E
Passive participation -E, D -E, -E
With the reward and punishment mechanism
Company A/Company B Active participation Passive participation
Active participation Ph+D, Ph+D q+D, Q-E
Passive participation Q-E, q+D PL-E, PL-E
Fig. 2 The overall architecture of our proposed ORPGT model
Environmental Science and Pollution Research
1 3
Level 2: The appearance is flawless; and there are traces of
use but no dents.
Level 3: There are some small scratches, slight dents but ,
no bending.
Level 4: A few apparent scratches, no bending.
Level 5: Many scratches, but no bending.
Level 6: The display is damaged and aging.
The research was conducted using the Weka1 tool that
employs fourmachine learning algorithms. These algorithms
are readily available machine learning algorithms used in
predicting standard pricing. Ultimately, they are used to find
the best features and are to be used alongside the Gaussian
Processes, the Linear Regression, the Random Forest Tree,
and the Multiplayer Perception to predict the recycling price
of these mobile phones.
The Gaussian Process and Linear Regression results are
similar, and theircorrelation coefficients and R2 values are
close to 1. The Multiplayer Perception’s correlation coeffi-
cient is 0.9444, which shows poor performance. The correla-
tion coefficient of Random Forest Tree algorithm is 0.9949,
which is the optimal algorithm among the four Fig.3.
The database is maintained in Excel 2016, and the all
the above-mentioned machine learning algorithms were
executed with Weka tool. Table5 gives the time needed
(second) for building the model (TM) for each algorithm
and their mean square error (MSE), root mean squared error
(RMSE), linear prediction function (LPF), linear predic-
tion R2(LPR), relative absolute error (RAE), root relative
squared error (RRSE) and correlation coefficient (CC). The
abbreviations of four algorithms areGaussian Processes
(GP), the Linear Regression (LR), the Random Forest Tree
(RFT), and the Multilayer Perceptron (MP).
Meta‑heuristic algorithm tofind thebest‑balanced
prices forthemarket
Meta-heuristic algorithms have many applications in dif-
ferent fields. The game theory in concert with meta-heu-
ristic algorithms will improve the efficiency of accurately
Fig. 3 Scatter chart between actual price and predicted price for different algorithms
1 Weka 3: An open-source machine learning tool is based on Java
and released by GNU. Data mining and visualization are done with it.
Environmental Science and Pollution Research
1 3
calculating the best-balanced prices for the market. After
analyzing the four different types of machine learning
algorithms, the results indicate that the efficiency of
therandom forest tree algorithm is comparative giv-
ing better results.Hence, the meta-heuristic algorithm
is based on thegametheoryand the random forest tree
algorithm to find the Nash Equilibrium by using the pre-
dicted major parameter, the depreciation coefficient (c).
For the experiment featured in this paper, 13 attrib-
utes about the phone (from the database) are taken as
the input to calculate the depreciation coefficient (c)
of the phone. Some of the trade in sites also give the
prediction of the recycling prize by comparing them
with the current model and the latest rate. This prize
is referred as the “standard payout” which helps to
maximize enterprise profit of or minimize consumer
loss. The model in Fig.4shows a simple idea of how
both enterprise and the consumers are satisfied with
maximum profit and minimum loss, respectively.
As shown in Fig.4, every enterprise has a different
cluster of customers. And the quantity of phones that
could be recovered from each enterprise will influ-
ence recycle prize by its own recovery price provided
for consumers along with the competitor’s recovery
price provided for consumers. For example, the recov-
ery prize offered by the enterprise’s is relatively high
for their customers when compared to competitors as
the enterprise gets new business from customers when
they exchange devices in the same market. Hence, this
section explains the utility function commonly used
by both the enterprise and the consumer. Every group
consists of one enterprise and a cluster of consumers,
and we define the utility function as Ue −Uc. The
objective is to find the recovery price sets for each
enterprise in the market which could let the total util-
ity function get the maximum value. The result is not
only to maximize profits for enterprise, but also to
have a higher consumer satisfaction with the recovery
price. The total utility function of the whole market
is the sum of each group’s utility function. Simula-
tion is conducted by Weka with the original data from
the market. The traditional methods give the results
exponentially, and after several trial and error proce-
dures, three of the meta-heuristic algorithms: Tabu
Search (TS), Variable Neighborhood Search (VNS),
and a hybrid of the Genetic Algorithm (GA) and VNS
are utilized to find an acceptable solution.
Tabu Search (TS)
Tabu Search(TS) is one of many gradient search techniques
designed for large, combinational optimization problems.
Tabu Search uses an individual serial search method. With
the taboo table, the search trajectory and the search direc-
tion can be recorded and tracked to prevent “premature con-
vergence” (Gmira etal. (2021)). Tabu Search gives better
results with both 2 or 3 permutations that are used. For this
research, 2-opt and 3-opt neighborhood structures have been
used. Under such structures, large-scale instances would
have good scaling capabilities and robustness (Žulj etal.
(2018)).
Variable Neighborhood Search (VNS)
Variable Neighborhood Search (VNS) is another meta-
heuristic framework that is used for solving (combination)
optimization problems. Variable Neighborhood Search
is one of the popular local search algorithms for different
kinds of neighborhood structures to search alternately, which
Table 5 Result comparison for
four available algorithms Algorithm TM MSE RMSE LPF LPR RAE RASE CC
GP 15.24 235.3553 309.2562 y = 0.9766x + 68.586 0.9768 13.81% 15.22% 0.9883
LR 0.93 234.0691 307.7538 y = 0.9782x + 62.691 0.977 13.73% 15.15% 0.9885
RFT 0.34 143.601 206.2108 y = 0.9804x + 55.462 0.9898 8.42% 10.15% 0.9949
MP 134.88 327.368 668.6814 y = 0.8997x + 313.78 0.8918 19.20% 32.92% 0.9444
Fig. 4 Relationship between customers and enterprises with different
recovery prices for different models
Environmental Science and Pollution Research
1 3
achieves a balance between intensification and diversifica-
tion. VNS identifies what local optimal solutions that are
likely to be found in a neighborhood, but not necessarily
in a neighboring neighborhood (Hansen etal. (2019) and
Djenić etal. (2016)). Moreover, the global optimal solution
is the local optimal solution of all possible neighborhoods.
As TS was implemented with 2 neighbor structures, VNS
was executed though four neighbor structures, i.e., 2-opt,
3-opt, 4-opt, and the insertion.
Genetic Algorithm andHybrid Meta-heuristic
Algorithm(GA andHGA)
In order to enhance the efficiency of results, the hybrid meta-
heuristic algorithm may be a more effective optimization in
comparison. Considering the steady performance of GA and
VNS, they are hybridized to form a hybrid meta-heuristic
algorithm for finding better solutions for the optimal recy-
cling price of second-hand mobile phones. Ultimately, the
idea of using this hybrid method is to optimize the total
utility function better. Fig.5 shows thee permutation: 3-OPT
and 4-OPT, respectively, which is used in the Tabu Search
and other proposed meta-heuristicalgorithms.
The algorithm HGA is the hybrid genetic algorithm
used for finding the optimal recycle price for the second-
hand mobile phones and is given below in Fig.6 to explain
the procedure of the proposed hybrid method byusing the
Genetic Algorithm(GA)and the Variable Neighborhood
Search (VNS).
Experimental results anddiscussion
Problem definition throughquestionnaire
This work mainly focuses on using game theory for get-
ting ORP. As a part of the preliminary work, we ran-
domly distributed 265 online structured self-administered
questionnaires. The respondents come from different prov-
inces across the country, covering distinct age groups and
using a wide range of mobile phone brands. SPSS2 tool is
used to analyze the survey data. 71.7% (190 out of 265) of
the survey respondents lacked clarity on the recycling price.
Among them, 71.6% (136 out of 190) will adopt negative
recycling behaviors such as postponing or abandoning recy-
cling. In the remaining 28.3% of the survey respondents who
understand the recycling price, nearly half believe that the
recycling price is unreasonable. While continuing to analyze
negative consumers, 56.6% (150 out of 265) of the survey
respondents left their used mobile phones at home. Moreo-
ver, 81.8% of the respondents expressed a willingness to
recycle that increased as the price of recycling increases.
Based on the initial survey, we understand that the price of
second-hand mobile phones is unintelligible for consum-
ers, and the lower price limits the recovery rate of mobile
phones.
Identify theoptimal recycling prize
The following parameters are used to predict the recycling
price, and the mathematical functions are also mentioned
in Table6.
Recycling old mobile phones is the first phase of research
to maintain customer benefit and the enterprises’ profits.
When considering the enterprise, the amount spent to
buy back used devices is an important factor in overall
profitability.
For the enterprise:
Equation (1) represents the expense of a phone that the
second-hand recycling company should pay.
Equation (2) represents the total number of recycled
mobile phones. It is calculated by the recovery rate and the
number of second-hand phones available on the market.
(1)
A
=
(
1+a
1)cD
(2)
M=eH
Fig. 5 Neighborhood of structure permutations
2 SPSS®:A data analysis software from IBM®
Environmental Science and Pollution Research
1 3
Fig. 6 The code of HGA
Environmental Science and Pollution Research
1 3
Equation (3) represents the enterprise’s expenses during
the recovery phase, and it equals the product of the recov-
ery price per cell phone and the number of cell phones
recovered.
Phase 2: Refurbish and sell.
After buying the second-hand phone, the profit or little
margin is fixed by refurbishing the used phone and selling
them back on the market as previously owned.
Equation (4) represents the total revenue, which is the
product of the total number of recycled mobile phones and
a second-hand phone sale price.
Equation (5) represents expenses during the refurbish-
ment and resale phase; it includes the cost per mobile phone
in the refurbishment phase and the tax on mobile phones
sold.
Equation (6) represents the enterprise’s utility function,
which is the total revenue minus the expenses incurred in
both the recovery phase and the refurbishment phase.
(3)
F1
=AM
(
1+a
1)ecDH
(4)
C=NM =NeH
(5)
F2=BM +a2NM =BeH +a2NeH
(6)
Ue
=CF
1
F
1
=eH
[(
1a
2)
NB
]
a
1ecDH
(7)
Uc=DE
Equation (7) represents the utility function of the con-
sumer, which is the price when the customer buys the phone
minus the recycling price received by customers.
As mentioned above, our approach uses this model to iden-
tify the optimal recycling prize (ORP) through which both
the customers and enterprises benefit. As per the requirement
of the algorithms, the parameter setting and the assumptions
made are listed in Table7 below:
According to Li’s (2018) definition of the recovery channel
model, for ith enterprise, its recovery rate ei depends on its own
E value and its competitors’ E value. Thus, the recovery rate is
For each group, the amount of second-hand mobile phones
that each enterprise could recover from the customers are
evenly distributed; however, this may be influenced by its
recovery rate ei. The Mi is defined as follows:
(8)
e
i=ex
iNi+yix
n
i,k1
NiN
k
Table 6 The set of parameter Parameter Description
a1Tax on recycling used mobile phones
a2Tax rate for sale after the mobile phones refurbished
c Mobile phone depreciation coefficient
e Recovery rate
A The company recovers the expenditure of a second-hand mobile phone
D The price when the customer buys the phone
M Number of recycled mobile phones
H Number of second-hand mobile phones in the market
C Total revenue
N Sale price of a second-hand mobile phone after refurbishment
B The cost of refurbishing a used mobile phone
E Recycling price received by customers
F1Expenses during the recovery phase
F2Expenses during the refurbishment and resale phase
Ue The utility function of the enterprise
Uc The utility function of the consumer
Table 7 The assumption of parameter
Parameter Value
a10.02
a20.03
e 0.75
xi0.25 × 104 ± 0.10 × 104
yi0.30 × 104 ± 0.10 × 104
zi0.045 ± 0.010
H 10000
ETN 0.93
n 50
Environmental Science and Pollution Research
1 3
The recovery price that enterprises gives to the customer
and the enterprise’s selling price is defined as
Each enterprise’s cost depends on the price difference
between the second-hand mobile phone recycling and selling
price, which is related by zi. The Bi is defined as follows:
(9)
(10)
Ni=ziE
Overall the utility function for the enterprise model can
be defined as:
The meta-heuristic algorithms’ goal is to maximize the
UTotal ; Equation (14) defines the maximum value as
We randomly selected 20 well-known standard mobile
phones as samples to test the meta-heuristic algorithms
and compared them to the current state of device recy-
cling. The initialization of the parameters is shown in
Table8 above.
From Fig.7 above, it’s apparent that the traditional total
utility values are lower than each total utility value calcu-
lated by the meta-heuristic algorithms by using the game
theory model. In this paper, the Nash Equilibrium truly
could let customers and enterprises reap more benefits com-
pared to the traditional price. As per our analysis, each one
of the three algorithms has its advantages when compared
with one another depending on the hardware. In the param-
eter settings of the model because of hardware constrains,
(11)
Bi
=z
i(
1+a
i)(
N
i
cD
)
(12)
Uei
=CF1F2=NeH
(
1+a2
)
AeH Z
[
(1+a2)N
i
A
]
(13)
Uci =DEi
(14)
U
total =
n
i=0
Uei U
ci
Table 8 Parameters for meta-heuristic algorithms
Parameter Value
Tabu search
Tabu list length 50
Number of neighborhood structure 2
Stopping criteria 1500
Variable neighborhood search
Number of neighborhood structure 4
Stopping criteria 1500
Hybrid genetic algorithm
Population size 20
Mutation rate
1
population size
0.8
Crossover rate 0~0.8
Generation 10
Hybrid VNS stopping criteria 100
Fig. 7 Results compared with TS, VNS, and HGA
Environmental Science and Pollution Research
1 3
the epochs for the VNS hybrid in the GA are only 100, while
the epochs for individual TS and VNS are 1500. Hence, if
the parameters for HGA are higher, the results of the HGA
may even give us better results. As per observed results, we
analyzed the issues found with different attributes. However,
we believe that the traditional price can be comparatively
enhanced with theproposed model.
Visualization ofrole oftheattributes withrecycle
price
Fig.8a shows that the recovery price based on different
attributes has been employed for the study. Generally, the
price of every brand depends on the following attributes
and also affects the price of recycling price. In the People’s
Republic of China, Apple brand mobile phones are particu-
larly coveted for their special features. As such, the particu-
lar features of every phone including its brand name are con-
sidered important factors. Furthermore, the recycling price
of 4G mobile phones is significantly higher than that of 5G
devices. For data collection, 5G technology is only used for
high-end mobile phones from Huawei and Apple. It is only
promoted in developed areas because the secondary market
has apparent limitations. 4G technology is the central mobile
phone signal technology currently in circulation, with a
broader audience under necessary factors such as the target
population and application scope. Therefore, the recycling
price will be affected by the applicable population, the
communication technology’s maturity, and available areas.
Fig.8b shows that the recovery price of an Apple iPhone’s is
much higher than that of other brands. However, the market
shares occupied by Apple and Huawei may change at any
given time; thus, this does not have a direct effect on the
price of recycling a device. Fig.8c shows the data is that of a
normal distribution tendency. The mobile phone chip version
directly affects the recycling price, which is more explicit for
Huawei and more implicit for Apple. Fig.8d shows data of a
centrally distributed tendency, and the RAM (random access
memory) sizes of devices directly impact the recycling price,
which is opposite to that of the purchase price. Fig.8e shows
that for all samples, the recycling price of phones with 32
GB ROM (read only memory) is significantly lower than
the standard, while others remain stable and concentrated.
The ROM size does not have a regular distribution on the
recovery price; hence it has no direct effect on the price of
recycling a device. Fig.8c shows the data is that of a nor-
mal distribution tendency. The chip version directly affects
the recycling price, which is more explicit for Huawei and
more implicit for Apple. Fig.8d shows data of a centrally
distributed tendency, and the RAM (random access memory)
sizes of devices directly impact the recycling price, which
is opposite to that of the purchase price. Fig.8e shows that
for all samples, the recycling price of phones with 32 GB
ROM (read only memory) is significantly lower than the
Fig. 8 Recovery price based on some of the attributes like a mobile mobile communication technology, b company, c version, d RAM, e ROM,
f CPU
Environmental Science and Pollution Research
1 3
standard, while others remain stable and concentrated. The
ROM size does not have a regular distribution on the recov-
ery price; hence it has no direct impact. In Fig.8f, different
CPUs (central processing unit) show a sinusoidal distribu-
tion tendency, and the overall data shows an upward trend.
Therefore, phones with higher-performing CPUs have a
greater recycling value. CPU types have a direct effect on
the recovery price of phones.
For Fig.9a, data presents a sinusoidal distribution and
reaches a peak between the interval [200, 500], and the over-
all downward trend is evident. The recycling price of sec-
ond-hand mobile phones drops significantly over time. Time
difference plays the most explicit role among all attributes.
In Fig.9b, the level is defined by the degree of complete-
ness of the appearance of the second-hand recycled mobile
phone. However, the distribution of each unit is uniform
and consistent. Hence, the impact of the level on the recov-
ery price can be ignored. Fig.9c shows the two types of
operating systems (OS) involved, IOS (Apple) and EMUI
(Huawei), and only discuss the most suitable OS of each
sample appearance settings of OS. The IOS operating system
shows a paragraph-like upward trend along the time axis,
while the EMUI operating system displays a growth wave
curve. Since OSs are updated with the timeline, the longer
the second-hand phones go before they are sold, the lower
the optimal OS version of the device. Fig.9d shows that
screen size (inches) has a centrally distributed trend, peak-
ing at 6.5 inches. The edge data is 4.7 inches and 7.7 inches,
in which the recycling price has an explicit downward
tendency. Hence, the intermediate size is more comfortable
to be accepted by customers. In Fig.9e, the data distribution
has no apparent regularity. First, the recycling prices of dif-
ferent companies are partially concentrated. Second, when
the Screen resolution is 24 dpi and 17 dpi, the data is con-
centrated and stable, but when the screen resolution is 22 dpi
and 19 dpi, the recycling price of used mobile phones drops
significantly. Overall data, the fluctuation range is implicit,
so the mobile phone’s resolution level slightly affects the
recycling price.
In conclusion, obtained by data visualization analysis
through Weka. Among all attributes, time difference has
the most explicit impact. It affects recovery prices by influ-
encing other co-factors. The lower the time difference of
second-hand mobile phones, the higher the recycling value
for consumers.
Data visualization of the data set compares the optimal
recovery price derived from the above mathematical model
and meta-heuristic. Time difference is the most important
factor. The impact of other attributes on the recovery price
is also the same. It can be demonstrated that the model and
derivation mechanism established in this paper are in line
with the actual situation. Cohen etal. (2021) has proposed
a model for solid waste recycling to identify difficulties that
a municipality and its residents face in building and operat-
ing infrastructure for recycling under the extended producer
responsibility law. Similar study will be carried out even
for market and enterprise resale prize based on the local
governing body.
Fig. 9 Recycle price based on the attributes like a time difference, b level, c operating system (OS), d screen size, e screen resolution
Environmental Science and Pollution Research
1 3
Conclusion
In the analysis of consumer recycling behavior, the question-
naire survey shows that more than 80% of consumers cannot
obtain the recycling price of second-hand mobile phones.
And more than half of them will passively participate in
mobile phone recycling due to that fact. This research has
processed and analyzed thousands of second-hand mobile
phone recycling price data to make the recycling price of
second-hand mobile phones more transparent to consumers.
This research establishes a second-hand mobile phone recy-
cling price model based on mobile communications technol-
ogy, company, phone name, version, RAM, ROM, time dif-
ference, CPU, operating system (OS), purchase price, screen
size (inches), screen resolution, and level. The data model
shows that the recovery time difference significantly impacts
the recovery price. This data model reduces the lag of con-
sumers’ knowledge of the recycling price and minimizes the
possibility of their passive participation. Furthermore, con-
sumers could realize the dramatic impact of recycling time
difference on recycling prices through this model. This may
then encourage consumers to participate in mobile phone
recycling earlier and more actively.
However, this model only collects four mobile phone
brands with the largest share in the Chinese market: Apple
and Huawei, Xiaomi, and Vivo. Data from other brands such
as Samsung and ZTE are missing. In the future, research
may focus on comprehensive analysis (that includes other
major brands such as ZTE and Samsung.) As mobile phone
series are updated quickly and often, future research should
analyze more comprehensive and updated mobile phone
brand data. And moreover, we would also implement the
Shekarian and Flapper (2021) closed loop analysis in future
when this paper referring to important configurations of the
circular economy (CE) has received considerable attention
in sustainability matters. It is composed of characteristics
that, when identified, studied, and categorized, help not only
to a better understanding of the current contributions in the
literature but also lead to formulating new models.
For recyclers, the study analyzes the behavior of recy-
clers through Hawk-Dove game and prisoner’s dilemma. The
Hawk-Dove game shows that recyclers are in a dominant
position in the market, and they determine the recycling
price of used mobile phones. Since second-hand mobile
phone recycling requires cost borne by the recyclers them-
selves through personal rationality and benefit maximiza-
tion, recyclers tend to provide consumers with lower prices.
The recycling of these devices requires the participation of
a great deal of mobile phone recyclers. Prisoner’s dilemma
shows that some recyclers’ passive participation will lead to
other recyclers’ inactive participation. Therefore, the gov-
ernment should participate in the second-hand mobile phone
recycling market and regulate recyclers. When recyclers
are passively involved in recycling, the government should
urge recyclers to increase taxes or face fines as punishment
mechanisms.
When consumers are actively engaged in the process,
the government should encourage recyclers by lower-
ing taxes or appropriating funds. This study uses meta-
heuristics to analyze the Nash Equilibrium between
consumers and recyclers in the commercial behavior
of recycling mobile devices.
The results shows a tendency that to reach the maxi-
mum profit in the whole market, the recyclers should
provide higher prices compared to the traditional prices
for the second-hand mobile phone
From the data visualization by using machine learning
algorithm in Weka, it’s clear that the time difference is
the core factor that affects the fluctuation of the recov-
ery price.
However, this research has some limitations: (1) the
second-hand mobile phone recycling price model
contains only four mobile phone brands with the larg-
est share in the Chinese market: Apple and Huawei,
Xiaomi, and Vivo; data from other brands such as Sam-
sung and ZTE are missing.
The feasibility, necessity of the reward, and punish-
ment mechanism in the process are only discussed at
the theoretical level.
Due to the random factors in the meta-heuristic, the
algorithm has the probability to diverge from the local
optimal solution and shift to a more global optimal
solution. Therefore, fixed input cannot guarantee uni-
fied output.
With the above stated limitation of the mathematical
model, some phones whose optimal price estimated by
the meta-heuristic algorithm is higher than their new
price.
Future research may focus on building a more completed
mathematical model based on comprehensive second-hand
mobile phone recycling data and take account the higher
importance of the time difference factor. As mobile phone
series are updated quickly, future research can analyze more
comprehensive and updated mobile phone brand data. More
practical rewards and punishment are needed to regulate the
behavior of recyclers. As Zhang etal. (2020) has mentioned,
the game theory model is built here to find out how effective
funding is for this recovery and recycling system in China.
So the researcher will be exploring more in WEEE.
Supplementary Information The online version contains supplemen-
tary material available at https:// doi. org/ 10. 1007/ s11356- 021- 17061-w .
Environmental Science and Pollution Research
1 3
Acknowledgements We are thankful to the leaders and the manage-
ment of the institution.
Author contribution Author contributions: Dr Sujatha and Dr Ken-
nedy are the supervisors of this research and guided the scholars. Liu
Zixuan designed the mathematical model and algorithms.Yihang Liu
collected the dataset, Yang Qiu and Zixuan Liu worked in the software.
Yihang Liu and Wanying Dou edited the whole document and initial
study was carried out by themand Zixuan Liu. All authors contributed
to the study conception and design.
Funding This project is supported by the Institute for Societal and
Contemporary computing by Wenzhou-Kean University.
Data availability Code is available; it can be given through git hub
link after acceptance.
Declarations
Consent for publication Not applicable
Competing interests The authors decl are no competing interests.
References
Arı E, Yılmaz V (2016a) A proposed structural model for house-
wives’ recy- cling behavior: a case study from Turkey. Ecol Econ
129:132–142. https:// doi. org/ 10. 1016/j. ecole con. 2016. 06. 002
Arı E, Yılmaz V (2016b) A proposed structural model for housewives
recycling behavior: a case study from Turkey. Ecol Econ 129:132–
142. https:// doi. org/ 10. 1016/j. ecole con. 2016. 06. 002
Chi X, Streicher-Porte M, Wang MY, Reuter MA (2011) Informal
electronic waste recycling: a sector review with special focus on
China. Waste Manag 31(4):731–742. https:// doi. org/ 10. 1016/j.
wasman. 2010. 11. 006
Cohen C, Halfon E, Schwartz M (2021) Trust between municipality
and residents: a game-theory model for municipal solid-waste
recycling efficiency. Waste Manag 127:30–36
Deng WJ, Giesy JP, So CS, Zheng HL (2017) End-of-life (EoL) mobile
phone management in Hong Kong households. J Environ Manag
200:22–28. https:// doi. org/ 10. 1016/j. jenvm an. 2017. 05. 056
Djenić A, Radojičić N, Marić M, Mladenović M (2016) Parallel VNS
for bus terminal location problem. Appl Soft Comput 42:448–458
Du J, Fu H, Han J, Ren X (2014) Obstacles and countermeasures of
waste cell phone recycling modern economic information. Mod-
ern Econ Inform pp 195–196
Forti V, Balde CP, Kuehr R, Bel G (2020) The global E-waste monitor
2020: quantities, flows and the circular economy potential. In:
United Nations University/United Nations Institute for Training
and Research. International Telecommunication Union and Inter-
national Solid Waste Association, Bonn, Geneva and Rotterdam
Gmira M, Gendreau M, Lodi A, Potvin JY (2021) Tabu search for the
time-dependent vehicle routing problem with time windows on a
road network. Eur J Oper Res 288(1):129–140
Hadzic S, Mumtaz S, Rodriguez J (2013) Cooperative game theory
and its application in localization algorithms. In: Hanappi H,
etal. (eds) Game Theory Relaunched, vol 8, pp 173–187, https://
doi. org/ 10. 5772/ 2563, URL https:// books. google. com/ books?
hl= zh- CN& lr= & id= Ad6dD wAAQB AJ& oi= fnd& pg= PA173
& dq= Coope rative+ Game+ Theory+ and+ Its+ Appli cation+
in+ Local izati on+ Algor ithms & ots= sV- CBhvn7- & sig= BsZuJ
U5aDN 32uWU 3LGFj xNzZ6 KI#v= onepa ge&q= Coope rative%
20Game% 20The ory% 20and% 20Its% 20App licat ion% 20in% 20Loc
aliza tion% 20Alg orith ms&f= false
Hansen P, Mladenović N, Brimberg J, Pérez JAM (2019) Vari-
able neighborhood search. In: In Handbook of metaheuristics.
Springer, Cham, pp 57–97
Hicks C, Dietmar R, Eugster M (2005) The recycling and disposal of
electrical and electronic waste in China—legislative and market
responses. Environ Impact Assess Rev 25(5):459–471. https:// doi.
org/ 10. 1016/j. eiar. 2005. 04. 007
Li S (2018) Research on the recycling pricing strategy of waste elec-
tronic products based on game theory URL http:// gb. overs ea.
cnki. net/ KCMS/ detail/ detail. aspx? filen ame= 10187 90578. nh&
dbcode= CMFD& dbname= CMFDR EF
Liu Y, Su X (2019) Research on pricing of waste electrical and elec-
tronic equipment fund based on game theory. Game Theory J
Green Sci Technol pp 259–264, https:// doi. org/ 10. 3969/j. issn.
1674- 9944. 2019. 18. 092
Liu H, Huang T, Lei M (2013) Dual-channel recycling model of waste
electrical and electronic equipment and research on effects of gov-
ernment subsidy Chinese. J Manag Sci 21:123–131
Liu J, Gan Z, Cui H (2019) Conception of mobile phone recycling plat-
form based on block chain technology. Resource Recycl 12:34–36
Myerson RB (2013) Game theory: analysis of conflict. Harvard Uni-
versity Press, Harvard
Needhidasan S, Samuel M, Chidambaram R (2014) Electronic
waste – an emerging threat to the environment of urban India.
J Environ Health Sci Eng 12(1):36. https:// doi. org/ 10. 1186/
2052- 336x- 12- 36
Shekarian E (2020) A review of factors affecting closed-loop supply
chain models. J Clean Prod 253:119823
Shekarian E, Flapper SD (2021) Analyzing the structure of closed-loop
supply Chains: a game theory perspective. Sustainability 13:1397.
https:// doi. org/ 10. 3390/ su130 31397
Shi Y, Wang J, Zhang Z (2020) Pricing strategy based on game theory
for two-stage reverse supply. Chain Ind Eng J 23:45
Smith JM, Price GR (1973) The logic of animal conflict. Nature
246(5427):15–18. https:// doi. org/ 10. 1038/ 24601 5a0
Su C (2018) Analysis on the influencing factors of consumers’ used
mobile phone recycling willingness.
Tan Q, Duan H, Liu L, Yang J, Li J (2018) Rethinking residential con-
sumers’ behavior in discarding obsolete mobile phones in China.
J Clean Prod 195:1228–1236. https:// doi. org/ 10. 1016/j. jclep ro.
2018. 05. 244
Wang M, Li Y, Shi W, Hao L, Quan S (2020) The dynamic analysis
and simulation of different government regulations in product-
recovery supply chain Systems Engineering -Theory & Practice.
Syst Eng Theory Pract 40:103–118
Wei Y, Zhu Y, Zhuang Y, Long G (2018) A recycling system of waste
mobile phone based on supply chain – taking nanning city as an
example tax paying. Tax Paying pp 149–151
Yin J, Gao Y, Xu H (2014) Survey and analysis of consumers’ behav-
iour of waste mobile phone recycling in China. J Clean Prod
65:517–525. https:// doi. org/ 10. 1016/j. jclep ro. 2013. 10. 006
Ylä-Mella J, Keiski RL, Pongrácz E (2015) Electronic waste recovery
in Finland: consumers’ perceptions towards recycling and re-use
of mobile phones. Waste Manag 45:374–384. https:// doi. org/ 10.
1016/j. wasman. 2015. 02. 031
Zhang B (2019) State transformation goes nuclear: Chinese national
nuclear companies’ expansion into Europe. https:// doi. org/ 10.
1080/ 01436 597. 2019. 16271 89
Zhang Y, Jiang G, Yang G (2019) Emotional attachment, income effect
and recycling willingness of consumers’ used mobile phones: an
Environmental Science and Pollution Research
1 3
empirical analysis based on. Endowment Effect Theory J Econ
Perspect 35:193–199
Zhang D, Cao Y, Wang Y, Ding G (2020) Operational effectiveness of
funding for waste electrical and electronic equipment disposal in
China: an analysis based on game theory.Resour Conserv Recycl
152:104514
Žulj I, Kramer S, Schneider M (2018) A hybrid of adaptive large neigh-
borhood search and tabu search for the order-batching problem.
Eur J Oper Res 264(2):653–664
Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Closed-loop supply chains (CLSCs) are seen as one of the circular economy’s leading approaches for reducing our natural environment load. Many CLSC models require collaboration among different parties. Game theory (GT) offers a way to consider the profits of all parties in a CLSC, providing insight into the costs and benefits to the involved parties in an objective and quantitative way. Presently, available reviews on the use of GT, in the context of CLSC, are quite limited and consider only a few relevant elements. Here, we present a new and more extensive framework, focusing on the collaboration structure of CLSCs. It contains a content-based analysis of 230 papers based on a four-step systematic literature review process. The characteristics studied are channels for collection, reprocessing and selling, the planning horizon, and the types of games. The structures found are graphically reviewed, leading to 196 different structures. The results show that, so far, most attention has been paid to the dual-channel collection, where collection by two retailers (dual-retailer) is the most studied case. With respect to selling, most attention has been paid to situations with two selling channels (dual-selling), i.e., one channel managed by a manufacturer and one channel managed by a remanufacturer. Studies have prioritized the role of manufacturers as that of the leader and collector. Finally, a number of directions for further research are pointed out.
Article
Full-text available
Closed-loop supply chain (CLSC) as one of the important configurations of the circular economy (CE) has received considerable attention in sustainability matters. It is composed of characteristics that, when identified, studied, and categorized, help not only to a better understanding of the current contributions in the literature but also lead to formulating new models. This research presents one of the first in-depth studies to investigate factors influencing CLSCs. It concerns the investigation of the models which are designed based on the game theory (GT). Therefore, the reviewed works focused on cooperation and competition among the game participants. A systematic literature review is implemented as a four-step process consisting of material collection, a descriptive analysis, category selection, and evaluation stage to review and discuss the works that focus on CLSC and use GT simultaneously. A content-based analysis is carried for the final works, which include 215 papers. The identified characteristics of these papers are classified into 12 main categories. Moreover, they are divided into subcategories to highlight the contribution of each paper. Accordingly, results are derived, and gaps are explained.
Article
Efficient and sustainable municipal solid-waste management is a social, environmental, and economic challenge. One practice that enhances the success of municipal solid-waste management is building and operating a household separation at source infrastructure for recycling. In this study, we use game-theory tools to identify difficulties that a municipality and its residents face in building and operating infrastructure for recycling under the Extended Producer Responsibility law. The model presents a holistic and broad perspective on the social and economic parameters that affect the efficiency of recycling in a municipality. We explain the strategies available to players and the factors affecting the utilities gained by waste separation and recycling. We present several Nash equilibria in pure and mixed strategies and specify the coordination game conditions. The model identifies parameters and their effect on the decision to recycle. The study presents changes needed to streamline a recycling system for an efficient equilibrium.
Article
This paper analyzes the characteristics of waste electrical and electronic equipment (WEEE) market in China and the findings are as follows: WEEE are still seen as valuable goods, and a recovery and recycling system is thus in place naturally in China like in many other developing countries. A game theory model is therefore built here to find out how effective funding is for this recovery and recycling system. Analysis of this model reveals: (1) Some dismantling enterprises (formal dismantling enterprises) are in this funding system, while others (informal ones) are not, thus forming a dual WEEE dismantling system, after the funding system was launched; (2) To ensure effectiveness of the funding system, the funding provided for formal enterprises for each piece of WEEE they handle, after deducting the incremental cost incurred for environmental compliance, must exceed the difference between the unit selling prices of the dismantled devices at formal and informal dismantling enterprises. This difference is RMB 70.2, RMB 67.9, RMB 22.2, RMB 79.4 and RMB 68.3 respectively for televisions, refrigerators, washing machines, air conditioners and computers; (3) When the funding system was launched, WEEE recovery enterprises would raise WEEE prices to seek profit and this would channel part of the funding into the recovery segment of the market.
Article
Most of the literature on state transformation focuses on China’s relations with African, Asian and Latin American countries and the National Oil Companies’ overseas expansion to show that China has become fragmented, decentralised and internationalised. This article contributes novel findings by focusing on China’s relations with Europe and the actions of China’s National Nuclear Companies (NNCs). It shows that NNCs, which have become relatively autonomous actors, often pursue their agendas of expansion into Europe without much coordination with, or even in contradiction to, other ministries’ agendas and interests, especially the Ministry of Foreign Affairs. Instead of being orchestrated by the central government, their expansion reflects considerable disorganisation and sometimes undermines China’s official strategy. The article demonstrates this through case studies of NNCs’ involvement in the UK and Romania.
Article
Significant progress has been achieved in China's electronic waste (e-waste) management, since a series of laws and regulations based on extended producer responsibility began to be enforced in 2011. In 2016, China's second batch of e-waste catalogue, which includes the mobile phone, was given priority. This study intended to propose potential approaches for addressing obsolete mobile phones management by examining residents' returning and recycling preferences and awareness in a typical city – Foshan, China, via face-to-face questionnaire surveys. The residents expressed their keen awareness of potential hazards caused by mobile phones and actively supported collection activities. However, 62.1% of residents stored their obsolete mobile phones at home, while only 4.7% of the mobile phones ended up in regulated treatment enterprises. The results indicated that most residents had much higher expectation on benefits from their obsolete mobile phones than their actual value, although only 1/3 of them declared the benefits would hindered their participation in collection activities. The formal collection channels, the convenience of collection facilities and guarantee of information security would also accelerate the collected amount. Additionally, this study investigated the structure of collection system and the relevant flow of mobile phones, and shed light on implications towards future studies and managerial implementation.
Article
Given a set of customer orders and a routing policy, the goal of the order-batching problem (OBP) is to group customer orders to picking orders (batches) such that the total length of all tours through a rectangular warehouse is minimized. Because order picking is considered the most labor-intensive process in warehousing, effectively batching customer orders can result in considerable savings. The OBP is NP-hard if the number of orders per batch is greater than two, and the exact solution methods proposed in the literature are not able to consistently solve larger instances. To address larger instances, we develop a metaheuristic hybrid based on adaptive large neighborhood search and tabu search, called ALNS/TS. In numerical studies, we conduct an extensive comparison of ALNS/TS to all previously published OBP methods that have used standard benchmark sets to investigate their performance. ALNS/TS outperforms all comparison methods with respect to both average solution quality and run-time. Compared to the state-of-the-art, ALNS/TS shows the clearest advantages on the larger instances of the existing benchmark sets, which assume a higher number of customer orders and higher capacities of the picking device. Finally, ALNS/TS is able to solve newly generated large-scale instances with up to 600 customer orders and six articles per customer order with reasonable run-times and convincing scaling behavior and robustness.
Article
A questionnaire survey and interviews were conducted in households and end-of-life (EoL) mobile phone business centres in Hong Kong. Widespread Internet use, combined with the rapid evolution of modern social networks, has resulted in the more rapid obsolescence of mobile phones, and thus a tremendous increase in the number of obsolete phones. In 2013, the volume of EoL mobile phones generated in Hong Kong totalled at least 330 tonnes, and the amount is rising. Approximately 80% of electronic waste is exported to Africa and developing countries such as mainland China or Pakistan for recycling. However, the material flow of the large number of obsolete phones generated by the territory's households remains unclear. Hence, the flow of EoL mobile phones in those households was analysed, with the average lifespan of a mobile phone in Hong Kong found to be just under two years (nearly 23 months). Most EoL mobile phones are transferred to mainland China for disposal. Current recycling methods are neither environmentally friendly nor sustainable, with serious implications for the environment and human health. The results of this analysis provide useful information for planning the collection system and facilities needed in Hong Kong and mainland China to better manage EoL mobile phones in the future.
Article
This paper considers the Bus Terminal Location Problem (BTLP) which incorporates characteristics of both the p-median and maximal covering problems. We propose a parallel variable neighborhood search algorithm (PVNS) for solving BTLP. Improved local search, based on efficient neighborhood interchange, is used for the p-median problem, and is combined with a reduced neighborhood size for the maximal covering part of the problem. The proposed parallel algorithm is compared with its non-parallel version. Parallelization yielded significant time improvement in function of the processor core count. Computational results show that PVNS improves all existing results from the literature, while using significantly less time. New larger instances, based on rl instances from the TSP library, are introduced and computational results for those new instances are reported.