A Study of the Japanese Seru-s Manufacturing System

Master Thesis for MSc BA - Operations & Supply Chains

Profiles: Production & Distribution, and Services

A Study of the Japanese Seru-s Manufacturing System

by Shuo Sun S1579908

First Supervisor: Dr. J.A.C (Jos) Bokhorst Co-assessor: Prof. dr. ir. J. Slomp

University of Groningen Faculty of Economics and Business

Landleven 5 9747 AD Groningen

November 2011

Acknowledgement

I would like to start by thanking my supervisors Dr. J.A.C. Bokhorst and Prof. dr. ir. J. Slomp.

First and foremost, I am very grateful to Jos Bokhorst for his excellent guidance from the beginning to the completion of this thesis. He offered me valuable ideas, suggestions and criticisms with his profound knowledge in research. Especially at the beginning, when I made very slow progress due to difficulties in getting used to differences in both culture and language in Groningen, he showed amazing patience and kindness, encouraging me to overcome all these obstacles. I would like to acknowledge Professor Jannes Slomp, as my thesis co-assessor, he provides a lot of valuable feedbacks and supports. Both my supervisor and co-assessor have played important roles in training my research thinking and English writing. Without their help and guidance, it would not have been possible for me to finish this thesis. Jos Bokhorst and Jannes Slomp, thank you very much, I have been lucky to be your Master student.

I would like also to acknowledge W. van Brussel, R. van Nikkelen Kuijper, D. Duis and J.

Faber who coordinated and participated in my interviews at Philips Drachten.

Special thanks to Zhao Zhao, Yue Han, Hristo Alexandrov and Victor Akerblom, they provided a lot of help during my student life in Groningen.

Last but not the least, I would like to express my absolute deepest love to my parents, especially to my father. All their unconditional love and support will be embalmed by me forever. I love you all.

November 2011 Shuo Sun

Groningen, the Netherlands

Abstract

Seru-Seisan is a Japanese manufacturing system which was invented at the beginning of 1990s. Many Japanese manufacturing companies in the field of electronics, such as Canon and Sony have been successfully implementing this system. However, it is not popular among manufacturing organizations and academic researchers beyond Japan. This study elaborates on the existing English studies of Seru-s. It provides information about many aspects of the Japanese Seru-s and compares the Seru-s and cellular manufacturing (CM) by means of a literature review. The conclusion is: they are exactly the same. To empirically verify this conclusion, a company was visited and several interviews were conducted. In the Seru-s literature, ‘Yatai’ is a Seru-s where all technical and managerial tasks are handled by a single cross-trained laborer, and Yatai is the ideal final stage of assembly (Seru-s’ evolution) for companies. However, Yatai is not regarded as the ideal final stage in assembly in the case.

Key Words: Seru-s (Seru-Seisan), cellular manufacturing (CM), conveyor lines, custom-lines, mass production, evolutionary trajectory/path

Table of Contents

Section 1 Introduction...

Section 2 Theoretical Background………..

2.1 The Definitions of Seru-s ………... 2

2.2 Seru-s’ Characteristics: Kanketsu, Majime and Jiritsu………....4

2.3 The Derivation of Seru-s……….5

2.4 What are the Reasons for the Emergence of Seru-s?...6

2.5 The Advantages and Disadvantages of Seru-s Compared with the Conveyor Lines….….9 2.6 The Evolution of Seru-s………..12

Section 3 Research Methods………...

Section 4 Literature Review on Seru-s and CM...

4.1 The Differences between Seru-s and Cellular Manufacturing from the Previous Studies.14 4.2 Seru-s and CM, Are They Really Different from Each Other?...19

4.2.1 The arguments are based on the differences between Seru-s and CM in table 1…………...19

4.2.2 The arguments are based on the evolutionary perspective………...…………21

Section 5 Case: The Assembly Systems of Shavers at Philips Drachten...

5.1 A Brief Introduction to Royal Philips Electronics and Philips Drachten……….. 27

5.2 The Discussions about the Old/New Assembly Systems of Shavers at Philips Drachten.28 5.2.1 The old assembly system for shavers………....…...28

5.2.2 The new assembly system for shavers (Thor)……….…….29

5.2.3 The driving forces of the conversion………...29

5.2.4 The philosophies behind the old and new assembly system, and Philips culture………...31

5.2.5 The essential factors to ensure the smooth operation of the new assembly system……...32

5.2.6 The advantages and disadvantages of the new assembly system………....…....33

5.3 First Stage of Seru-s/CM? And What is the Evolutionary Path of Shavers’ Assembly System in Philips Drachten?...34

5.4 Yatai: Ideal End Stage in Assembly Evolution for Philips Drachten?...34

Section 6 Conclusions, Limitations and Further Recommendations...

References………..

Appendix………

Section 1 Introduction

From the early 1980s, the field of operations management has been affected a lot by Japanese production management (Schonberger, 2007). There are many well known elements of production management that have been developed by Japanese scholars or manufacturers, for example, Kanban, Jidoka, 5S etc. In 1992, a new innovative manufacturing system was introduced in Japan and the name of the system is called Seru-Seisan¹. This new system has been replacing the traditional conveyor line gradually over time in some Japanese manufacturing fields, especially in the electronics field (Yin et al, 2006). A lot of Japanese electronics corporations have applied or they are willing to apply Seru-s for their assembly processes. For example, the famous Japanese electronics manufacturers Hitachi, NEC, Panasonic, Fujitsu, Sony and Canon, already abandoned their traditional long I-shape conveyor lines and employed Seru-s (Kono, 2004a; Iwamuro, 2002; Yoshita and Kimura, 2004; Kataoka and Yamada, 2001; Yagyu, 2003; Noguchi, 2003; Tsuru and Isa, 2002;

Takeuchi, 2006). In other words, Seru-s philosophy for manufacturing has been widely accepted by Japanese electronics industry since the early 1990s and the high performance feature of Seru-s has been confirmed and approved, such as the high flexibility of Seru-s system.

C.G. Liu et al (2010) argue Seru-s has attracted considerable interests from both academic research and production practice in Japan. However, there are two constraints that prevent Seru-s to be internationally promoted and implemented. First, Seru-s is a relative new production management system in the fields of manufacturing, operation and industrial engineering. Consequently, there are only a few people, academic researchers and managers of manufacturing industry who are familiar with the philosophy of Seru-s out of Japan.

According to our review, the number of published English-written papers on Seru-s is less than ten. Most of the papers that deal with this progressive manufacturing organization are written in Japanese. Therefore, international academic researchers or industrial managers could not gain sufficient information about Seru-s. Second, most researchers who study Seru-s are in Japan, so a considerable amount of studies on Seru-s have been done only in Japan. Overall, although a lot of studies have been done on Seru-s and Japanese scholars have emphasized the outstanding performance of this manufacturing organization, there is still a need for international scholars to access those available results on Seru-s, so that it might trigger and facilitate further research on this innovative system. Moreover, Yin et al (2008) mention ‘cell’ is the English word for ‘Seru’. Therefore, Seru-s can be directly translated to English word ‘Cell Production’ or ‘Cellular Manufacturing (CM)’. A question can be raised:

“Is the concept of Seru-s new in field of manufacturing or it is just the same as or similar to CM with a different name?” The concepts of Seru-s and CM share many characteristics, so there are still a lot of people who think they are exactly the same concept but with different names. In this paper, the above question is going to be addressed. In addition, the evolutionary trajectory of Seru-s is studied and presented by Yin and colleagues (Yin et al, 2008). The ideal

1 ‘Seru’ is a Japanese word and its English translation is ‘cell’, ‘Seisan’ is a Chinese word in Japanese pronunciation which means production or manufacturing. In the following of this paper, I will employ the word

‘Seru-s’ instead of ‘Seru Seisan’ in order to apply the consistent and convenient terminology to develop my discussion.

final stage of Seru-s’ evolution is called perfect Seru-s or ‘Yatai’ in their study. Yatai is the Japanese word for ‘street stall’ where one single person prepares and sells fast food. In this case, Yatai means one single laborer Seru-s, which all technical and managerial tasks will be handled by only one cross-trained laborer within a seru. In another major study, C.G. Liu et al (2010) argue Yatai should not be considered as the final modality in Seru-s’ evolution.

However, they did not give the detailed reasons behind their argument. The ambiguous explanation about Yatai might not be the ideal final stage of Seru-s’ evolution needs further clarification.

This research therefore mainly attempts to address the above problems by reviewing current available studies on Seru-s and also by conducting a descriptive case. Yin et al (2008) and C.G.

Liu et al (2010) have already discussed some aspects of Seru-s in English version. In this study, their findings will be represented and extended in a more systematic and concrete way, and make further contributions to clarify and develop Seru-s system. Thus, the main aim of this study is:

By reviewing current studies on Seru-s, to introduce and disseminate the knowledge about Seru-s in order to increase people’s familiarity of this innovative manufacturing system.

Before achieving the above research aim, it is crucial to have a good insight into the context of Seru-s. Therefore, the remainder of the paper is organized as follows: Section 2 conducts the theoretical background of Seru-s to gain more knowledge on it and to achieve our research aim in particular. This section includes many aspects and perspectives about Seru-s which are discussed from different points of view, these aspects about Seru-s are: definitions, features of Seru-s, derivation, reasons for the emergence, advantages & disadvantages, and types of Seru-s along its evolutionary trajectory. Research aim is refined to focus more on the comparison between Seru-s and cellular manufacturing (CM) at the end of section 2. Section 3 describes the research methods of this paper. Section 4 conducts literature review to compare the differences between Seru-s and CM, and to discuss whether or not Seru-s is a new manufacturing organization compared to CM. Section 5 deals with a descriptive case: the current assembly process of shavers at Philips Drachten. In this section, we will firstly give a brief introduction of Philips Drachten and its old and new assembly systems. The original research aim will be also associated with the case study to some extent. Moreover, based on the discussion about the Seru-s’ evolutionary trajectory in section 2, the discussion will be further lead into the case to detect whether or not there is a fixed evolutionary trajectory of the shavers’ assembly process at Philips Drachten. The conclusions, limitations and further recommendations of this paper are given in section 6.

Section 2 Theoretical Background 2.1 The Definitions of Seru-

There are many definitions of Seru-s from different scholars, and some accepted definitions are as follows:

 “Seru-s is a manufacturing organization (in most cases, a manufacturing assembly organization) which consists of one or several workers, and all assembly tasks of a product are completed within the Seru-s” (Yamada and Kataoka 2001: p. 73).

 “Removing conveyor lines, Seru-s is a jiritsu² manufacturing unit in which all required tasks of producing a product are completed by one or several workers who manage several equipment of different type” (Shinohara, 1995).

 “A highly jiritsu organization that completes work orders from start-to-final and is managed by one or several workers” (Iwamuro, 2002: p27).

 “Seru-s is a method of production carried out by a limited number of workers by removing conveyor line.” (Kono, 2004b).

 “Cell (Seru-s) production exemplifies the shift toward human-centered production systems. A cell-production system is a production system in which a single worker or small team of production workers (two to five members) perform multiple production jobs (multitasking) in short segment lines. Cells are, with few exceptions, arranged in U-shaped lines in which unfinished components enter at a point adjacent to the point that they leave as finished products” (Isa and Tsuru, 2002: p. 550).

 “A seru-s is a manufacturing line (an assembly line in most cases), which consists of equipment and people that are dedicated to one or several products; and which presents the characteristics: kanketsu, majime, jiritsu.” (Yin et al, 2008-E01: p12).

Yin et al (2008) also state Seru-s should employ a line layout. However, there is a questionable standpoint in their definition for Seru-s, if the perfect Seru-s ‘Yatai’ is a single worker who performs all the assembly tasks, obviously a single Yatai does not seem like a line layout. Therefore, a question could be raised: why Seru-s should only employ a line layout and are there any another possible layouts for Seru-s? The authors also present the purpose to develop the definition is to distinguish Seru-s from other manufacturing organizations.

However, the authors did not provide additional information on the above issue to their Seru-s’ definition. The above question should be answered in order to develop a distinct definition of Seru-s. In this paper, the research questions have been established; due to the fact that the above question is beyond the scope of this research, the related information concern the issue is not going to be provided. Therefore, the definition of Seru-s given by Yin and colleagues can be improved if the ‘layout’ issue can be taken into account. Overall, the contents of those definitions are almost same to a certain degree; they are overlapped at certain point of view. From my understanding, “Seru-s is a manufacturing system which completely replaces the traditional conveyor line for the assembly processes, and the assembly processes are handled by one or several high-skilled laborers who are fully supported by all the possible resources (tools, equipment, co-workers etc).”

2 Jiritsu is a Japanese word; the meaning of Jiritsu, Kanketsu and Majime are explained in the later part of this study.

2.2 Seru-s’ Characteristics: Kanketsu, Majime and Jiritsu

To recall the definition of Seru-s from Yin et al (2008), Seru-s presents three main characteristics named as kanketsu, majime, jiritsu and these three features formed the backbone of Seru-s. Yin et al (2008) also give a detailed discussion about these three features.

In this paper, the three features will be explained briefly to establish a general idea for the readers.

Kanketsu: it means that all necessary assembly tasks are managed and completed from the start-to-finish within the Seru-s (Yin et al, 2008). The equipment and the cross-trained laborers act as the two main resources in Seru-s. All the necessary resources for performing the assembly tasks should be well prepared beforehand within the assembly Seru-s in order to achieve Kanketsu. Tools and equipment should be placed close proximity to the laborers, enabling them to perform both routine and customized assembly tasks (Isa and Tsuru, 2002).

Generally speaking, assembly tasks are manual works in an assembly Seru, so only simple equipment such as hand tools are needed to perform the manual works. Because the relatively simple equipment is normally self-made with low costs and easily duplicates to each Seru-s, the equipment issue is not often a problem in assembly Seru-s (Yin et al, 2008; Yoshita, 2004;

Shinohara, 1995; Isa and Tsuru, 2002). However, there is a need of continuous cross-training of laborers to perform successfully in Seru-s, because it is difficult for a laborer to become fully cross-trained; huge amount of time, energy and money should be invested in by both employees and corporations to continuously improve the laborers’ skills. Therefore, Yin et al (2008) argue the continuous cross-training of workers is prerequisite to achieve Kanketsu for Seru-s.

Majime: In Seru-s, the concept of Majime means minimizing the physical space between the adjacent workstations or minimizing the distance between the adjacent laborers. The main purpose of Majime is to smooth the continuous processes in the result of the quick material flows and frequent communications. The achievements of Majime can be: reduction of the physical inventory, shortened throughput time, facilitation the information exchange and almost zero WIP (work-in-process) inventory in a Seru-s. Yin et al (2008) state Majime can be treated as the supplement of Kanketsu, because Kanketsu needs all the necessary resources be available within a Seru-s, and Majime needs all the unnecessary resources have to be removed from a Seru-s in order to ensure the continuous processes go smoothly. The Majime of an assembly Seru can be easily achieved by employing cross-trained laborers (Yin et al, 2008).

Jiritsu: It means autonomous and self-managed, learning and evolutionary (Yin et al, 2008).

Autonomous and self managed define Seru-s as an independent system, and there is no need of the external control for a Seru-s operation, because laborers also play a managerial role in the Seru-s that they are working on. Yin et al (2008) present two types of tasks from the managerial perspective: the routine tasks (designing laborers’ schedules, communicating with other companies, allocating laborers into different positions along with a specific Seru and so on) and emergent tasks (to manage the broken equipment, to deal with the urgent orders and so on). All these tasks should be addressed within Seru-s due to the effect of Jiritsu. Moreover,

the learning aspect of Jiritsu allows Seru-s to create organizational knowledge and it is the first step for Seru-s to be evolvable; this first step can be done by the effective communication between the laborers and the continuous cross-training. The continuous improvements of laborers’ skills and experiences are the essentials of the Seru-s evolution, and to make all laborers being fully cross-trained is the final goal that companies pursued. Yin et al (2008) also define the ‘organizational evolution’ as a process that for an organization autonomously and dynamically changes its architecture and behavior following the knowledge creation. And they also divide organizational evolution into two types: Adaptive-evolution and Self-evolution. Both adaptive- and self- evolution will be discussed further in sections 4.

To summarize, Kanketsu, Majime and Jiritsu are the three main features of Seru-s and they formed the backbone of Seru-s system. If any one of them is absent, the structure of the high-performance Seru-s might collapse immediately (Yin et al, 2008).

2.3 The Derivation of Seru-s

In 1992, many short assembly lines replaced the long I-shape assembly line to assemble one specific video camera in a factory of Sony (Takeuchi 2006; Kimura et al, 2004; Nonaka et al, 2004). Each single camera was assembled on a short line as on the traditional long I-shape conveyor. Figure 1 shows an example of how a long conveyor line is decomposed.

Figure 1: An example of a long I-shape conveyor line is decomposed into short assembly lines

Due to the example of figure 1, the long I-shape conveyor line with eight operational stations (OP) is dismantled into four short assembly lines, and each short line consists of four operational stations. The low-skilled laborers that perform the routine assembling tasks on the traditional long conveyor line are replaced by the cross-trained workers. In each short line, an autonomous group contains certain number of high-skilled workers who take the complete responsibility for assembling one or several specific products, and job rotation and training are addressed within the autonomous groups. For instance, in order to assemble type 1 product (labeled as output 1 in figure 1), two laborers (W1 and W2) work on the first short line. There are four assembly operations (OP1, OP4, OP6, OP8) should be performed in order

to finalized type 1 product. From the above diagram we can see that, W2 performs or controls three operations which are OP4, OP6 and OP8. However, W1 is only performing on OP1. In other words, W2 is higher skilled than W1, therefore, W2 can handle more operational stations during the assembly process than W1 to finalize type 1 product. With continuous development and innovation involved, the manufacturing lines become shorter and shorter;

and the system could be adjusted quickly in order to meet the unstable market demands.

Abandoning of the long conveyor lines and adoption of short assembly lines brought considerable benefits to Sony (C.G. Liu et al, 2010; Yin et al, 2008).

Since 1992, soon after the influence of Sony’s high performance due to the implementation of the new innovative manufacturing system, a considerable number of manufacturers discomposed their long conveyor assembly lines into the new layout. According to Yin et al (2006), Kon Tatsuyoshi, an employee at Sony named the new manufacturing system (figure 1) as Seru-s³ in 1994. Over time, Seru-s has become one of the most popular and important manufacturing system in Japan. It is especially a well-known manufacturing concept in final assembly processes of Japanese electrical and electronics industries (Yin et al, 2008).

Furthermore, the new transformed layout for the assembly processes in figure 1 is called divisional Seru-s. The further information about the divisional Seru-s will be presented in section 2.6.

2.4 What are the Reasons for the Emergence of Seru-s?

In 20^th century (from 1920s to earlier 1990s), the dominated and the most efficient production system was the mass production. The universally accepted definition of mass production is:

the production of large amounts of standardized products including and especially on assembly lines.⁴ Dertouzos et al (1989) and Prestowitz, Jr (1991) studied the success of Japanese industries in 1980s. Dertouzos (1989) finds that since the business environment stayed relatively stable in 1980s, Japanese corporations obtained many benefits by implementing long conveyor lines for mass production. In the electronics and electric industries, the low-end and highly reliable products emerged as a new business market and most manufacturing companies in Japan gained competitive advantages over their American counter parties by mass-producing these low-end products. High production volume but low production variety is the main characteristic of producing low-end electronic products.

However, the drawbacks of the long I-shape conveyor line have usually been overlooked by the managers of many manufacturing industries (Johnson, D. J, 2005), excess capacity, excess employment and excess debt are also formed a true portraiture for the traditional conveyor systems. One of the main disadvantages of mass production (long conveyor layout) is lack of flexibility, because it is hard to change production process or production design once the long conveyor line has been established. Bukchin et al (1997) discuss the challenges of mass production with the long conveyor lines. They find the layout of long conveyor use for manufacturing is not very good at addressing demands for varieties, design changes or customization.

3 The original name of Seru-s is Seru Seisan.

4 http://en.wikipedia.org/wiki/Mass_production

After 1991, the Japanese business environment changed unexpectedly. The low flexible conveyor line was abandoned and Seru-s emerged due to the following four main trends of the Japanese business environment (C.G. Liu et al, 2010; Yin et al, 2006; Yin et al, 2008).

Trend One: The Main Trend, Demand Pattern has Changed to High Variety and Low Volume (HVLV)

More and more diversified and customized products have been demanded on the Japanese market after the collapse of the economic bubble in 1991. There are three main reasons behind this: in Japan, the national income was negatively affected by the economic bubble burst, but the developments of various new technologies have never been stopped in the recent two decades. As a consequence, it was possible for people to shift their consumption from one relative expensive product to many cheaper substitutes in the domestic market due to the emergence of different technologies. Secondly, Japanese manufacturers were shifting their production from low-end products to high value-added products. After the collapse of the economic bubble, Japanese manufactures lost their market position gradually due to the stronger competitive pressure from other East-Asian countries (see this in ‘Trend two’), economic recession and the developments of new technologies. Consequently, Japanese manufacturers were forced to focus more on innovating and producing high-end products as well as adding more products’ ranges into their production after 1991 in order to survive in the world market. Thirdly, the developments of new technologies have shortened the life-cycle of products, especially the electronic and electric products. Johnson (2005) argues the shortened product life cycle is a main reason for increasing the demands of diversified products. Overall, the demands for various products increased rapidly in Japan after the collapse of the economic bubble in 1991.

A new system for manufacturing or assembly has to be developed in order to increase the flexibility, reduce the lead times on new product development and allowing greater customization and variety of products. Because of this situation, Seru-s emerged.

Besides the Seru-s manufacturing, many scholars also suggested an alternative manufacturing organization to address the high variety/ low volume (HVLV) business environment: the lean manufacturing (J.Slomp, J.A.C Bokhorst and R. Germs 2009; J. Jina et al, 1997). However, some scholars argued lean manufacturing was not efficient due to the changes in business environment compared with Seru-s (C.G. Liu et al, 2010; Liker, 2004; Imaoka, 2005). In the light of Womack et al (1990) and Matthias (2007), Lean production is a management philosophy derived mostly from the Toyota Production System (TPS) and identified as ‘Lean”

only in the 1990s. The TPS has been widely implemented by various firms in the world and it is a milestone in the fields of operation and production management. However, Liker (2004) and C.G Liu et al (2010) argue that there is not sufficient evidence to prove that the efficiency of TPS in other industries could be as great as in the car industry. According to the argument by C.G. Liu and his colleagues, the components are large in size for the assembly in car industry, but the components are relatively small for the high-end products in electronics industry and they are easy to transport. Therefore, the TPS fails to realize the expected high efficiency for the electronics industry.

However, we do not see a clear causal connection in their argument. The argument is quite vague because it does not clearly explain why TPS is not efficient anymore in the electronics industry. In our opinion, even though there is still no clear conclusion whether Seru-s is based on TPS or not, but at least, Seru-s and TPS are not contradictory to each other and they have a strong interrelation. Many ideas of Seru-s are still based on TPS (such as elimination of production wastes by optimizing one-piece flows, assembly process should be close to demands/customizations in order to deliver exactly what consumers want in the right quantities, time and locations), this fact cannot be denied. Thus, Seru-s can be treated as a system that developed and is not isolated from TPS at all. Actually, we have to distinguish the following facts: both traditional conveyor line and Seru-s systems are based on TPS to a certain degree, the layout or the function of the conveyor line lack of flexibility compared with Seru-s in assembly or production process, not the TPS lacks of flexibility compared with Seru-s. As a system was developed from TPS, Seru-s emerged. Of course, we have to also beware of the short-comings of TPS in order to make the further improvements. For example, Camuffo and Weber (2011) study the recent Toyota crisis and they conclude that there are still many weak points of TPS and many steps should be taken to improve the system. Although there are some weaknesses of TPS, its principles will influence the modern managements continuously (Isa and Tsuru, 2002).

Trend Two: More Pressure Comes from International Rivals

Before the collapse the bubble economy in 1991, Japanese electronics firms enjoyed their unbeatable position in the international market by producing low- and middle- end products.

After obtaining the reputation and competitive advantages in the market, Japanese firms also successfully competed with the US manufacturers in the category of high-end products.

However, after the burst of the economic bubble, Japan suffered from the negative consequences of the recession. The Japanese firms began to lose their market shares.

Moreover, competitive pressure to the Japanese industries was enhanced due to the rise of the economy of some Eastern-Asian countries especially China and South Korea (Leggett and Wonacott, 2002). Powell (2002) and Jiang (2004) argue the laborer costs are relatively low in China which drives the prices on many products down in the world. The low prices of Chinese low- and middle-end products had generated significant impacts on many developed countries, and Japan is one of them. China has gradually taken over the Japanese market position in the international market, especially the market position of the electronic products.

Yin et al (2008) present an example for this issue: “In an investigation about the global market shares of sixteen high-tech electronic products, China emerged as top of eight out of the sixteen products. Japanese electronics industries were losing market shares to China.”

Furthermore, the Japanese currency rose rapidly in value against US dollar in early 1990s, the export volume of Japan declined sharply due to the appreciation of the Yen. Consequently, Japanese industries lose their market shares further in the global market. In order to survive in such fierce competition, Japanese manufacturers began to outsource their products or develop new manufacturing organizations (Willette et al, 2004)

Trend Three: Japan went to Long Term Economic Stagnation after the Collapse of the Economic Bubble

The mass production are normally capital intensive, so if the automation of a production system is high, the capital investment to the system will be high too. After 1991, the Japanese economy went to the long period stagnation or even recession. The stock market crashed, and many companies went bankrupt. A lot of manufacturing firms could not afford the additional high capital investment anymore for the maintenances or other related costs required by the high automation for the production system. Furthermore, the high automation systems can not satisfy the requirements of the unstable external demands anymore due to the lack of flexibility. A new low cost but more flexible system for manufacturing was urgently needed by many manufacturers.

Trend Four: the Enthusiasm of Employees for Working was lowered by the Collapse of the Economic Bubble

The burst of the economic bubble in 1991 generated negative effects not only on Japanese economy but also psychologically influenced employees. After 1991, Japanese economy experienced decline. Many companies dismissed their employees or cut their wages in order to reduce costs⁵. However, those policies hurt the welfares of employees and lowered the enthusiasm and loyalty to their formal jobs. Moreover, the workers who operated on the conveyor lines were gradually getting dissatisfied with their tedious jobs. C.G. Liu et al (2010) state that: “The high specialization, as well as the strict time on the conveyor line, results in a monotonous work environment. In such a circumstance, worker tolerance and enthusiasm decrease in the long term.” Therefore, the employees asked for a more human-oriented working environment, and some more flexible tasks on the production lines were required by the employees. During the economic recession, instead of conveyors, companies were seeking a new production system to cope with the fluctuate demands in order to maintain reasonable profits, so the benefits of the employees could be granted to some extent.

2.5 The Advantages and Disadvantages of Seru-s Compared with the Conveyor Lines

The purposes to discuss the advantages and disadvantages of Seru-s in this part are: on one hand, to find the incentives for companies to convert their traditional flow lines to Seru-s, and on the other hand, the drawbacks of Seru-s should not be overlooked by the companies which have the intention to convert their flow lines. In the light of literature review on the available papers about Seru-s, the above can be summarized into four main advantages and four main disadvantages in general. These will be discussed individually as follows:

The Advantages

 The high flexibility is the first main advantage of Seru-s. Miyake (2006) states that by increasing or decreasing the number of workers or workstations, Seru-s can be adopted to address the external fluctuating demands. Seru-s normally consists of several mobile workstations and cross-trained laborers, and it is easy to establish a seru. From our general knowledge, the positions of laborers are fixed along the conveyor lines for

5 Source: Japan peaks under Nakasone then the bubble economy and collapse, Japan 06 Modern History, Facts and Details.

production or assembly processes, and the outputs are normally also fixed along the conveyors. Under- or over-production normally occurs under such a production system (C.G. Liu et al., 2010; Smunt and William., 1985; Ray and Carnall., 1976). In comparison, it is obvious that Seru-s is more flexible to make the adjustments by changing the number of workstations or cross-trained laborers to cope with the rapid variations in demand than traditional conveyor lines.

 The second main advantage of Seru-s is shortening the throughput time. Miyake (2006) also states that, to design and install a seru could be done in a few weeks, and the setup time would be effectively reduced compare with the time to set up a conveyor line (may be several months). The distance between workstations is shortened in Seru-s, the amount of time for employees to move from one station to another is shorten too.⁶ Moreover, a defective component will only affect a specific Seru-s process, but such component may affect the whole process when a conveyor line is implemented. Overall, Seru-s can effectively reduce the throughput time compared with conveyors.

 Lowered inventory quantities and costs are the third advantage of Seru-s. According to C.G. Liu (2010), by converting the traditional conveyor line to Seru-s, both inventories for work-in-process (WIP) and finished goods are reduced dramatically. The production or assembly processes by adopting conveyor lines need a considerable amount of WIP inventories as buffers to smooth the working processes between stations. However, the role of the WIP inventories played on conveyor lines is eliminated under the philosophy of Seru-s, because it is easy for Seru-s to adjust the operating ranges by neatly positioned the multi-skilled laborers, so the blocks and obstacles during the processes can be eliminated. As mentioned previously, the volume of production is small in Seru-s to address the unstable demands, and the real demands determine the production quantity.

Therefore, only a small quantity of end products is stored under Seru-s to protect both own and suppliers’ production schedules from being shocked by sudden demands. By comparing with the inventory’s waste of end products, far more waste in the entire production or assembly processes or even the whole supply chain can be reduced by minimizing the final product inventory, if the production level is kept.

 The fourth main advantage of Seru-s is to improve the employees’ enthusiasm and loyalty to their jobs, and it also facilitates companies to establish their own cultures.

Recall section 2.4, conveyor lines are adopted for mass production or assembly processes which are essentially repetitive. After the reverse conversion, laborers can perform more flexible tasks rather than the high-repeatability routine tasks. The decentralized Seru-s empowers employees more rights to perform their works, and their talents can be easily explored. As the distances between workstations are shortened under Seru-s, employees are closer to each other for working which drives more communications between employees. A work team can be well established by the increased communication and the team’s performance will be improved substantially. Fiol and Lyles (1985) argue the

6 Sometimes it is necessary for an employee to move from one workstation to another in order to perform tasks.

decentralized working environment will facilitate the organizational learning to establish a distinct organizational culture. Under the philosophy of Seru-s, laborers are multi-skilled and cross-trained for both operations and managements. The HVLV demand pattern allows high flexibility in Seru-s system, and the communication level between employees is increased obviously after the reverse conversion. Due to the decentralization, more powers are allocated to employees for decision making. Therefore, we can conclude that Seru-s is convenient for a company to establish a distinct culture.

Furthermore, in the traditional Japanese corporate culture, the wages of employees are based on their qualifications and experiences rather than on performance. Therefore, there is a common phenomenon that some employees really have a good work performance but still get a lower and fixed payment, because they have not gained enough seniority yet in their jobs. However, the rigid conventional payment systems were changed gradually after the adoption of Seru-s. According to Yin et al (2008), the skill levels of laborers who are performing on Seru-s are divided into four distinct classes by Canon: the Third-class, the Second-class, the First-class and S-class. S-class is the highest class in the skill-ranking system which S means ‘super person’. Kimura and Yoshita (2004) present an example for the efficiency of an S-class worker; they state that it only needs two hours for an S-class laborer to assemble a product which contains 2700 components. Companies generally offer considerable benefits to the S-class employees and it effectively encourages employees to work hard and learn from the new system continuously to develop their skills and performances in order to achieve the S-class.

Moreover, Miyake (2006) states that laborers can choose or design the elements of Seru-s such as workstations, tools, product carts by themselves to ensure the high efficiency of the equipment. Due to the high skill and the unique knowledge of Seru-s, laborers are also invited to participate in the layout designing of a new Seru-s or re-deploy an existing one (Olorunniwo and Udo, 2002). Furthermore, C.G. Liu et al (2010) also argue that, laborers are selected for the cross-training in order to work on Seru-s. As a result, it seems a lot of workers losing their jobs, because there are fewer laborers needed to work on a Seru-s after the conversion from the traditional conveyor line. However, laborers are normally re-allocated to the different departments rather than dismissed by their companies after the reverse conversion, so the total number of employees in a company is normally unchanged after the implementation of Seru-s (C.G.

Liu et al, 2010). Based on the above points, the loyalty and enthusiasm of employees to their jobs are increased substantially after the reverse conversion to Seru-s (Sakamaki, 2006).

The Disadvantages

 First of all, the expandability of a certain seru is limited. When product volume increases or decreases slightly, by adding or reducing number of workstations or number of serus to adjust the capacity, it is not difficult to address the problem. However, when there is a big fluctuation in product volume, for example, the product volume increases sharply to a certain level, the efficiency of Seru-s will decline dramatically and will be even lower

than the efficiency of conveyor lines. This is because, in general, there is a restricted number of laborers in a seru (up to fifteen); the additional workstations or serus will occupy too many areas when there is an unexpectedly increase in product volume. It is difficult and costly for Seru-s to adjust to cope with the large product volume (C.G Liu et al, 2010; Miyake, 2006; Sakamaki, 2006). Therefore, Seru-s lacks of expandability to some extent.

 Secondly, to train workers to be multi-skilled requires substantial investments. Kimura and Yoshita (2004) state a considerable amount of money and time have to be invested in laborers’ trainings, because multi-skilled laborers are the core of Seru-s, the cross-training is very important. In order to keep the skills most updated to cope with the unstable demands, the laborers’ cross-training should be continuous, so the long-term substantial investments for the cross-training have to be continuous too.

 Thirdly, there would be a high variable cost of production during the early period of Seru-s’ adoption. After the reverse conversion, many simple types of equipment replace the old heavy equipment used for conveyors, the simple type equipment is always designed and created by Seru-s laborers and engineers, and tools might also be specifically re-designed for Seru-s. Therefore, in the beginning period of Seru-s adoption, the variable cost of production might be high.

 Fourthly, laborers will face high pressure after the reverse conversion. Noguchi (2003) argues that some laborers are against the idea of their companies adopting Seru-s and such unwillingness has become an important management dilemma for the companies that want to implement Seru-s. The laborers are against Seru-s, because of the high pressure. Seru-s relies heavily on the laborers’ skills, in order to successfully implement Seru-s, selected laborers need to take several training courses across many disciplines, such as quality control, communication skills, production management, how to use the new equipment and tools and so on. The considerable amount of energy and time has to be invested by these laborers to learn the courses, communicate with other colleagues, develop skills, learn the operations of the new equipment and perform the formal tasks.

Higher pressure are pushed to these labors and some of them could be reluctant towards the Seru-s’ implementation. Normally, in a traditional conveyor system, laborers are assigned fewer tasks for assembling a small variety of products than in Seru-s and cross-training is not a prerequisite for working on the conveyors, so laborers are less pressured when performing on the traditional flow lines.

2.6 The Evolution of Seru-s

Figure 2 shows the evolutionary trajectory and stages of Seru-s. According to Yin et al (2008) and C.G. Liu et al (2010), the first stage of the evolution is the reverse conversion: converting from the traditional conveyor line to which is called divisional Seru-s and this conversion can be referred to in figure 1.

Divisional Seru-s

At the first stage of Seru-s’ evolution in figure 2 (see below), the long conveyor line is decomposed into many short lines and it is easy to convert. Laborers are cross-trained in some technical skills and none of managerial skills. These laborers are allocated separately to these short lines, and they also cooperate with each other to perform tasks as before. However, some laborers can perform multiple tasks and some of them can only perform one or two tasks. In general, the assembly processes be handled by each laborer are more than on the traditional flow line. Moreover, laborers only focus on technical tasks, and there is a supervisor to perform all managerial tasks, which the exact same as on the conveyor line.

According to Yin et al (2006), divisional Seru-s is more flexible than conveyor line in addressing business environmental changes. Furthermore, the expensive equipment used for conveyor lines will be replaced by self-designed and made equipment for divisional Seru-s, which is also cheap at the same time (Yoshita, 2004).

Rotating Seru-s

The second stage of Seru-s’ evolution is called rotating Seru-s. At this stage, laborers are fully cross-trained in technical skills to perform the tasks from start-to-finish for assembling one or several product types. It means these workers still lack managerial skills and knowledge, thus a supervisor is still needed to take responsibility of the managerial tasks. The rotating Seru-s is normally treated as a transitional stage by Japanese manufacturing industry for linking the first stage and the final stage of the evolution, which lead divisional Seru-s to the end stage. In figure 2, the rotating Seru-s presents as U-shaped. Assume there are three assembly operations for products with a fixed sequence, and all three laborers (W1, W2, W3) can perform all three operations individually. When he or she finishes a product, he or she will return to the first assembly operation for assembling a new product. Yin et al (2008) state that:

“The assembly operations are often performed on fixed stations, and laborers walk from one station to another to complete each assembly operation.” The big arrow in figure 2 presents the walking trajectory of laborers. Comparing with divisional Seru-s, rotating Seru-s is more flexible and efficient, because all laborers are fully cross- trained on technical skills.

Yatai, the Perfect Seru-s

In the light of Yin et al (2008), the end stage of Seru-s’ evolution is called ‘Yatai’ or the perfect Seru-s. In this stage, laborers are fully cross-trained in all technical and managerial skills. The idea is only one single laborer is needed to perform all the tasks (both technical and managerial) in a Seru. ‘Yatai’ is a Japanese word meaning ‘street stall’: one person cooks and sells fast foods on the street, and all the necessary steps for preparing the foods are managed and performed by him- or herself, there is no extra help from others. Figure 2 shows an example of three Yatai-s, and each Yatai is only managed by one fully cross-trained laborer. In this case, there is no bottleneck of the productivity from other laborers, each laborer on Yatai will be able to fully exploit his or her ability and talent. Moreover, like the rotating Seru-s, laborers who work on Yatai-s are also moving among the operations.

From the above discussion, Seru-s seems to be similar to cellular manufacturing, such as it can reduce the physical inventory level, minimize the buffers between working stations, increase in the autonomy of laborers etc. Therefore, the refined research aim is:

From both theoretical and practical perspectives, to discuss whether or not Seru-s is really a new creative concept compared to the Cellular Manufacturing.

Section 3 Research Methods

In this section, the research methods to compare Seru-s and CM will be established in order to achieve our refined research aim. Firstly, we are going to conduct a literature review of available articles on Seru-s. The literature will be gathered by using the search engines like EBSCO and Google Scholar. Due to the scarce English version literature of Seru-s, the references of available English-writing and some Japanese translated literature about Seru-s are checked and quoted. The contents of these referred articles will be carefully examined and selected in order to address the integrative analysis. Moreover, the related articles of CM and other fields are selected and reviewed to support the arguments of this paper. Secondly, a descriptive case is included in section 5. This case is about the assembly process of shavers at Philips Drachten. The aim of including the case is to explore the incentives of Philips Drachten for the conversion of its assembly system, and additionally verify whether or not

‘Yatai’ is the ideal end stage in Seru-s’ evolution to further support or reject our arguments on the comparison between CM and Seru-s from a practical case. This case is conducted by a company visit and interviews with three Philips’ specialists in three different positions. All of them have good knowledge about the assembly systems of Philips Drachten (one of them is a production manager, one is the engineer for cells/serus and one of them is responsible for engineering of the final assembly systems). In addition, one randomly chose operator who performs on the system at Philips Drachten was also interviewed. It is therefore possible to explore information from different angles by talking with different positioned employees.

Section 4 Literature Review on Seru-s and CM

4.1 The Differences between Seru-s and Cellular Manufacturing from the Previous Studies

In the light of J. Slomp et al (2002), a cellular manufacturing (CM) system is defined as: “the grouping of workers and machines into relatively independent cells, which are responsible for the complete manufacturing of a set of part types.” We can easily see that the definition of CM is very similar to the definitions of Seru-s which is provided in section 2.1. Moreover, as mentioned in section 1, Cell is the English translation for Seru and Cellular Manufacturing are the English words for Seru Seisan, which the abbreviation ‘Seru-s’ stands for in this paper.

Therefore, Seru-s can be directly translated to English word ‘cell production’ or ‘cellular manufacturing (CM)’. The concepts of Seru-s and CM are shared many keywords such as one-piece-flow, cross-training, job rotation, close located equipment and others (Yin et al, 2008), so there are still a lot of people thinking they are exactly the same concept but with different names. Yin et al (2008) and C.G. Liu (2010) argue that even though Seru-s and CM have the identical meaning literally, there are still many differences between the two concepts.

Moreover, C.G. Liu et al (2010) conclude the differences between Seru-s and CM from five perspectives: the theoretical background of formation, the practical applications, the equipment adopted for processing, the flexibility of layout’s adjustments and the evolutional trajectory.

Figure 2: The evolutionary trajectory and stages of Seru-s

These five perspectives are discussed as follows:

The first perspective, CM has a long history, and has been studied over several decades. The Group Technology (GT) is the foundation of CM; in other words, CM is an application of GT for production. Since the early 1970s, CM has become a famous concept in many fields such as manufacturing operations, production management, and industrial engineering etc. A considerable number of studies about CM have been published in journals and newspapers (Reisman et al, 1997; Yin et al, 2006). On the other hand, Seru-s is a relatively new concept with a short history. The term Seru-s was developed by Sony and named by a Sony employee in 1992. Nonaka et al (2004) and Kimura et al (2004) argue the theoretical foundation of Seru-s is the combination of Toyota Production System and Sony’s single-worker theory.

However, their arguments have not been well supported. C.G. Liu et al (2010) state that the theoretical foundation of Seru-s is still a source of debate in academia, and there is no final agreement at this moment. Furthermore, according to Yin et al (2008), most CMs are converted from job shops and serus are normally converted from conveyor lines. In order to distinguish the two conversions, Yin et al (2008) state that the conversion from traditional conveyor lines to Seru-s is called reverse conversion; and the conversion from job shops to CM is called traditional conversion due to its long history.

In a job shop, similar types of machines are allocated together as the workstation areas in a corner of the plant. Every single type of product has to be produced in a specific manufacturing planning, because each single product type has its own process path. However, under the philosophy of CM, similar products or product families with the associated machines are grouped together, so those products can be processed within the associated machines (Wemmerlov and Hyer, 1986). Miltenburg (2001) and Aase et al (2003) present that U-shape is the typical layout of CM for production which is unlike the layout of job shops, and machines are allocated close to each other in a sequence to manufacture the similar products or the product families. The process paths of the production can be coincided which is unlike the unique process path for each single product type in job shops. Moreover, according to Johnson and Wemmerlov (2004), Miltenburg and Zhang (1991), Job shops are easier to fit to new products than CM due to the high flexibility. A wild range of dissimilar products can be easily handled by job shops. On the other hand, similar products should be formed as a specific product family to ensure the sufficient volume for CM (it is based on the philosophy of Group Technology), normally, each single product type has a low volume and it is not optimal to manufacture single typed product individually under CM, otherwise the

product capacity will not be fully utilized. However, many scholars have argued that CM generally outperforms job shops. For example, Burbidge (1975) is one of the pioneers of GT and he has presented some advantages of CM over job shops such as better accountability, decreased setup time and lead time of production, more satisfaction of employees and so on (Yin et al 2008). Jonhson and Wemmerlov (1997) also analyze the substantial improvements of the performance by the traditional conversion. Furthermore, C.G. Liu et al (2010) state the main purpose of the traditional conversion is seeking for efficiency.

Now, let us consider the reverse conversion. As mentioned before, the term ‘reverse conversion’ was applied by Yin et al (2008) to distinguish the ‘traditional conversion’. The reverse conversion was used to describe the conversional process from traditional conveyor lines to Seru-s. Yin et al (2008) mention one reason that there are still no publications about the traditional conversion (to convert job shops to CM) in Japan. They state this is because most serus were converted from the traditional assembly lines in electrical and electronics industries, and these industries had excess capacities during 1990s (Yin et al, 2008). Most Japanese scholars argue Seru-s is a unique manufacturing system that was invented in Japan (Sony, 2005; Yamada and Kataoka, 2001; Noguchi, 2003). Due to the change of the business environment in 1990s, the aim of the reverse conversion is seeking for flexibility to address the demand fluctuation.

The last four perspectives about the differences between the two concepts from C.G. Liu et al (2010) will be discussed together as follows. The CM has been applied for both manufacturing and assembly processes in many industries. In most cases, Seru-s has been adopted for the assembly process in electronics and electric industries. Seru-s also does not need to apply the group technology, because the production variety in conveyor lines is low and production volume in Seru-s is low as well, so there is no need to allocate the similar products together to form product families after the reverse conversion to Seru-s. In general, it adjusts the layout of Seru-s to different shapes to cope with the demand changes; however, this can not be easily done in CM. In Seru-s, the resources for production are allocated close to each other with different shapes according to the requirements, but in CM, the heavy machines are allocated along with a specific cell, it is difficult to move the machines in different positions to adjust the layout. Moreover, according to Shinohara (1995) and our previous discussion, most serus are the assembly serus which only need simple equipment or tools for laborers to perform the assembly tasks, assembly serus rely heavily on the laborers skills. However, machines and equipment play an important role in CM because it also deals

with the manufacturing processes. The simple equipment be adopted by Seru-s is modified and redesigned from the heavy equipment used for traditional conveyors (Yoshita, 2004). The Seru-s laborers know better of the system than anybody else, so the necessary tools or equipment are normally self-designed by these laborers. However, equipment or machines are bought or designed specifically for the product families in CM, the initial capital costs of CM should be higher than Seru-s. Yin et al (2008) and C.G. Liu et al (2010) also state that, because the cross-trained laborers is the main resource for Seru-s implementation, Seru-s can continuously self-evolve and reach at a final stage which is called perfect Seru-s or ‘Yatai’.

However, there is no obvious evidence to support CM also can be self-evolved. C.G. Liu et al (2010) summarized the differences between Seru-s and CM from five perspectives. Besides those five perspectives, Yin et al (2008) discussed the differences in a more specific and detailed way. Table 1 summarizes the differences between the two concepts/systems according to the study of Yin et al (2008).

Table 1: Summary of the differences between Seru-s and CM by Yin et al (2008)

Seru-s CM

History Short (since 1992) Long (since 1920)

Origin TPS& Single laborer theory Group Technology

Conversion Reverse Traditional

Purpose of conversion Seeking for flexibility Seeking for efficiency Type of operation Assembly in most cases Manufacturing /Assembly Resource and capital costs Human-centered, Equipment

is less important, high initial capital costs

Both human and equipment are important, relatively low capital costs

Product variety For one Seru: low For one cell: high Single-laborer Yatai, exists in some cases Exists in some cases

Besides the above differences, Yin et al (2008) further investigate whether or not Seru-s and CM are similar manufacturing organizations. Hyer and Brown (1999) propose three critical linkages of a cell as: Information: information about the tasks is fully accessed by laborers under CM; Time: the transfer and waiting time (WT) between the tasks are minimized; Space:

all tasks perform on a specific cell should be physically close to each other. Furthermore, Wemmerlov and Hyer (2002) define a cell under CM from four aspects: Resource: human and machines are allocated to a set of similar products in a cell for processing. Spatial:

resources are grouped together to a cell and resources should be positioned as close as

possible to each other physically with a clear boundary. Transformation: all similar products can be transformed to a defined product family to share the process steps in a cell.

Organizational: a cell can be treated as a unit of administration in a company, for example, a cell can be seen as a control point, a cell is accountable for the performance improvement and so on. Yin et al (2008) argue these three critical linkages and four aspects of a cell are very important to CM and can be treated as the elements which are formed the backbone of CM. In order to investigate whether or not Seru-s and CM are the similar manufacturing systems, Yin and his colleagues further analyzed if or not the three characteristics (Kanketsu, Majime and Jiritsu) of Seru-s can be substituted by the three critical linkages or the four aspects of CM, and vice versa. According to the results of their study, all three critical linkages and the four aspects of CM can be achieved by Seru-s, in other words, all these three linkages and the four aspects of CM can be substituted by the three main characteristics of Seru-s. However, the three characteristics of Seru-s can also be substituted by these linkages and aspects of CM that were discussed above except that Seru-s is self-evolvable, but CM is not. To summarize the findings of Yin et al (2008), the only difference between Seru-s and CM is the evolutionary issue. Besides that, Seru-s and CM can completely work as the substitution to each other to achieve the same functions.

4.2 Seru-s and CM, Are They Really Different from Each Other?

4.2.1 The arguments are based on the differences between Seru-s and CM in table 1 In order to prove Seru-s and CM are the complete two different systems, there should be at least one unique characteristic of Seru-s that CM does not have and vice versa. In the other words, if there are some characteristics that can be shared and substituted between Seru-s and CM, then the argument of Seru-s is a unique new manufacturing system will not hold anymore to some extent. Most differences between the two systems in table 1 could become similarities or even the same issues to a certain degree. The following shows the discussions of a few viewpoints:

1. Even though Yin et al (2008) and C.G.Liu et al (2010) summarize some key differences between the Seru-s and CM, we still can not conclude that Seru-s is a new system which is different from CM. As Yin et al (2008) mention that, the previous organization types of CM and Seru-s should not be taken into account for considering the differences between CM and Seru-s. We can not control and decide the exogenous variables such as histories, origins, aims of the conversions.

2. The artificial terminologies such as ‘traditional conversion’ and ‘reverse conversion’ were

applied by Yin and his colleagues for distinguishing the two conversional processes, these terms do not exist in the related fields before their study. To recall the discussions in the previous sections, Seru-s is converted from conveyor line (reverse conversion) and CM is converted from job shop (traditional conversion), C.G. Liu et al (2010) and Yin et al (2008) argue this issue can be considered as a difference between Seru-s and CM.

However, there is also a circumstance that CM can be directly converted from conveyor line, for example, Johnson (2005) study the performance improvements due to the conversion from assembly flow lines to assembly cells in a real sheet-metal plant.

Therefore, converting from conveyor line is not the unique case that occurred by Seru-s, it also occurs for CM.

3. According to the differences of the operational types in the fifth row of the table, Yin et al (2008) argue that most Seru-s are designed to perform assembly tasks in the electronics and electric industries, and CM is implemented for performing both manufacturing and assembly tasks in a wide range of industries. However, Yin et al (2008) also state that, besides assembly serus, there are also machining serus and the utilization of the equipment is emphasized for machining serus. C.G Liu et al (2010) also present an example that Toyota has already successfully implemented machining serus for manufacturing. Thus, it seems that both of Seru-s and CM have been adopted by manufacturers for both manufacturing and assembly processes. In addition, both CM and Seru-s pay attention to the agility of labor force, e.g. the cross-training is emphasized (Yin et al 2008); Black (1991) also state that the workers in the cells are multi-process:

they are high skilled and can run more than one process, the different kinds of processes can be performed by these workers as well. Thus, we could say that the difference has been showed in the sixth row of the table 1 can be considered as the results of the different kinds of tasks are concerned (assembly tasks or manufacturing tasks); in other words, Seru-s is considered as human-centered with the less important role of equipment, because most serus has been classified as assembly serus, and the assembly tasks are normally the manual works which rely heavily on the skills of laborers. If Seru-s is classified as a machining seru, then the equipment is emphasized. And the required resources and capital investments will be indifferent between CM and Seru-s.

4. In the seventh row of table 1, Yin et al (2008) present the difference of product variety within a seru and a cell. As mentioned previously, a single seru normally is adopted for manufacturing or assembling one type product, but a cell is implemented for