The content and process of automation strategies

ARTICLE IN PRESS

Int. J. Production Economics 110 (2007) 25–38 www.elsevier.com/locate/ijpe

The content and process of automation strategies K. Sa¨fstena,, M. Winrotha, J. Stahrea,b a

Industrial Engineering and Management, School of Engineering, Jo¨nko¨ping University, Sweden b Division of Production Systems, Chalmers University of Technology, Sweden Available online 12 March 2007

Abstract If automation is to support the competitiveness for a manufacturing company, strategic as well as operational issues need consideration. To best support competitiveness, decisions concerning automation should be treated as one of several decisions in a manufacturing strategy. Furthermore, to fully utilize the advantages from automation, the manufacturing strategy content and process needs refinement. In this paper, improvement of the manufacturing strategy theory is suggested, mainly based on employment of human factors engineering. r 2007 Elsevier B.V. All rights reserved. Keywords: Semi-automated manufacturing systems; Automation strategies; Manufacturing strategy; Content and process; Human factors engineering

1. Introduction Considering the increasing global competition and the threats of e.g. outsourcing and off-shoring to low-cost countries, competitive manufacturing capability is a critical and urgent matter for manufacturing companies. Automated manufacturing systems are often regarded as highly efficient, potentially improving the competitiveness of manufacturing companies. In the manufacturing domain abundant literature addresses the concept of automation (e.g. Chang et al., 2005; Mehrabi et al., 2000; Yu et al., 2003). The literature deals with different types of technical solutions, such as advanced manufacturing technology (AMT), and Corresponding author. Department of Industrial Engineering and Management, School of Engineering, P.O. Box 1026, SE-551 11 Jo¨nko¨ping, Sweden. Tel.: +46 36 10 16 39; fax: +46 36 10 05 98. E-mail address: [email protected] (K. Sa¨fsten).

different ways of implementing automation. In the area of AMT, focus is on manufacturing process technologies that include for example computerized storage of information (Dean et al., 1992) or completely automated manufacturing solutions such as the SMART-cells (Fujimori, 1990; Makino and Yamafuji, 1988). The major problem with the technology oriented literature is that it focuses on the specific applications and the potential improvements but, unfortunately, fails to explain how to select technological investments that support a business (Hill, 2000). Many studies indicate that most automation decisions emanate from the top, and often the outcome is not what was expected when making the investment. When top management initiates automation, often with the aim to reduce manufacturing cost, the decision on automation tends to be the only concern, i.e. automation is the manufacturing strategy (Winroth et al., 2007). If the decision is pushed on the organization, without linkage to the

0925-5273/$ - see front matter r 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.ijpe.2007.02.027

ARTICLE IN PRESS 26

K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

manufacturing capabilities, such investments may become real failures. An automation decision formulated together, and in congruence, with the other decisions in a manufacturing strategy, i.e. automation strategy as part of the manufacturing strategy, has shown to be more successful (Boyer et al., 1996; Winroth et al., 2007). Within the manufacturing strategy domain, decisions concerning automation, as well as other decisions affecting manufacturing resources and providing competitive advantages through support of the overall strategic initiative of the firm, have been described (Maruchek et al., 1990). Level of technology is one of several decisions constituting the manufacturing strategy content (e.g. Skinner, 1969; Hill, 2000). Here, however, the question is automation or not, and the appropriateness of different levels of automation in different situations is not treated. This view of automation is also noticed by Sheridan (2002) as the ‘‘all-or-none fallacy’’, especially common among the non-technical public. When planning and implementing automated manufacturing systems, there are numerous issues to consider. In contrast to the process industries, systems in the manufacturing industry are rarely fully automated. A common solution is to integrate manual and automated operations into semi-automated manufacturing systems. Automation can involve automation of activities both at facilities level and on support systems level (Groover, 2001), i.e. physical issues as well as decision and control tasks can be automated (Frohm, et al., 2005). In order to fully utilize the capabilities of both humans and machines in a semi-automated manufacturing system, the interaction between them needs to be well conceived. Such interaction has traditionally been described in human factors engineering in the terms of function allocation, implying a system design process where functions are allocated to humans or to machines, respectively. The resulting function allocation may be described as the level of automation, ranging from entirely manual operations to full automation (Sheridan, 2002). Function allocation between human systems and technical systems is a far from trivial issue, and has been treated within areas where risks are extremely high, such as in aerospace and process industries (e.g. Sheridan, 2002; Inagaki, 2003). The applications of function allocation within the manufacturing industry are so far limited, although some initiatives have been undertaken (e.g. Fallon, 2001; Granell et al., 2006).

Automation decisions need to be made as a part of the other manufacturing strategy decisions. Existing models about the content and process of manufacturing strategy, emanating mainly from Skinner (1969) and Hayes and Wheelwright (1984), deal with automation very briefly as a question that is included in the process technology decision. The view on automation on a strategic level tends to be an ‘‘all-or-none’’ decision. To fully utilize the possible advantages of automation as supportive for manufacturing competitiveness, the strategic decisions concerning automation need to be linked to the operational issues of task allocation between human systems and technical systems, a problem addressed within the human factors engineering domain. This paper elaborates on the possibilities of refinement of manufacturing strategy content and process with support from the human factors engineering domain in order to improve the support from automation on manufacturing competitiveness.

2. Methods and materials Research presented in this paper was carried out as part of an ongoing Swedish research project, DYNAMO1—Dynamic Levels of Automation. DYNAMO is a 3-year project that ended in 2006. The DYNAMO project aims at dynamic levels of automation, i.e. a possibility to vary the level of automation according to the specific situation and the thereby associated requirements. DYNAMO is to provide industry with design, measurement, visualization, and management tools for dynamic levels of automation in manufacturing. Dynamic levels of automation are useful during multiple phases of the product realization process and are expected to increase manufacturing system’s overall robustness. This paper focuses on aspects mainly related to the management of dynamic levels of automation on a strategic level, i.e. formulation and use of automation strategies within the area of industrial manufacturing. The results presented in this paper are based on both theoretical and empirical material. 1 The project was financially supported by The Swedish Foundation for Strategic Research through its research program ProViking. A number of manufacturing companies also supported the research by actively taking part and giving access to their knowledge and premises.

ARTICLE IN PRESS K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

27

Table 1 Overview of completed case studies Case

Case Case Case Case Case

A B C D E

Industry

Size of company/site (no of employees)

Product: size, volume, complexity

Year of study

Telecom Automotive Wood Mechanical Furniture

1500 650 400 400 o100

Small, very high volume, complex Large, medium volume, complex Medium, medium, medium Small, high volume, simple Medium, medium, medium

2000 1998 2004 2004 2006

The empirical studies are carried out as case studies, mainly focusing on manufacturing system design and manufacturing strategy (Sa¨fsten, 1998, 2002; Bellgran and Sa¨fsten, 2005; Winroth, 2004). In total, five different case studies are presented in this paper, each of them representing different manufacturing companies. The size, the volume and the complexity of the products manufactured in the studied companies vary, as does the size of the companies; for an overview see Table 1. The studies have been performed by means of qualitative interviews, analysis of documents, and direct observations (Yin, 1994) during the period 1998–2006. The respondents have mainly been people involved in production issues, on different levels. On an overall level, the analysis of the empirical data from the case studies followed the procedure suggested by Miles and Huberman (1994), involving data reduction, data display, and conclusion drawing and verification. The rest of the paper is structured as follows. Initially, a review of the manufacturing strategy literature is presented, and automation on a strategic level is treated. Then automation on an operational level is outlined, followed by a description of some practical approaches to automation and their consequences. Thereafter the results are presented, and the paper ends with some conclusions and discussion. 3. Manufacturing strategy After 30 years of debate and discussion following Skinner’s seminal work on the importance of manufacturing (Skinner, 1969), a more or less common understanding today is the necessity of manufacturing to support the overall performance of a company. Managers are becoming increasingly aware of the competitive strength that manufactur-

ing can provide and everyone understands that manufacturing can be a competitive weapon (e.g. Roth and Miller, 1992; Hayes and Clark, 1995), at least if fit between manufacturing technology and business strategy is achieved (Kotha and Swamidass, 2000). Manufacturing is one of several functions that have to support the achievement of the overall objectives for a company. The task can be fulfilled with support from a well formulated and implemented manufacturing strategy since a manufacturing strategy comprises a series of decisions, which, over time, provide the necessary support for the relevant order-winners and order-qualifiers of the different market segments of a company (Hill, 2000). A strategy consists of the plan and the type of action needed to achieve defined objectives. Manufacturing strategy is here defined as a pattern of time- and market-specific decisions in structural and infrastructural areas supporting competitive priorities for a company. Manufacturing strategy is not only about making the correct decision supporting competitive priorities. According to Hayes and Pisano (1994) it is about creating and selecting operating capabilities for the future in a company. This is, however, embedded in the manufacturing strategy as such, since the decisions should reflect a longer perspective than the operational day-to-day decisions. A manufacturing strategy is a functional strategy, together with for example marketing, R&D, and accounting strategies. The functional strategies, in co-operation, support the business strategy of a company (Hayes and Wheelwright, 1984). Manufacturing strategy can be divided into strategy content and strategy process (Swink and Way, 1995). 3.1. Manufacturing strategy content The content of a manufacturing strategy concerns aspects such as manufacturing capabilities and

ARTICLE IN PRESS K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

28

strategic choices (Dangayach and Deshmukh, 2001). Manufacturing capabilities deal with the competitive priorities, which are achieved through a set of proper decisions within different decision areas, the strategic choices (see Table 2). The competitive priorities are most often listed as being cost, quality, different aspects of delivery, and flexibility (e.g. Wheelwright and Hayes, 1985; Ward et al., 1996; Hill, 2000). In manufacturing strategy literature the most commonly mentioned decision areas are production technology, capacity, facility, vertical integration, quality, production planning and control, workforce, and organization. More or less congruous views on the theme have been presented over the years (e.g. Hill, 2000; Skinner, 1969; Wheelwright and Hayes, 1985; Ward et al., 1996; Miltenburg, 2005). The structural decision areas are characterized by their long-term impact; they are difficult to reverse or undo and they often require a substantial capital investment (Hayes and Wheelwright, 1984; Wheelwright, 1984). The infrastructural decision areas are often considered to be more tactical in nature; they are built up by ongoing decisions and generally do not require extensive capital investment. It can, however, be quite costly to perform changes also among the infrastructural decisions, a consideration that should by no means be neglected. It is not a question of either infrastructural or structural issues; it is a question of the right combination. Automation, or similarly level of technology, is mainly treated within the structural decision area involving issues related to the production process. The decision areas described in the literature have different denominations for similar areas. The decision area Process design (Swink and Way, 1995) is also called plant and equipment (Skinner, Table 2 Manufacturing strategy content (Swink and Way, 1995; Dangayach and Deshmukh, 2001) Manufacturing strategy content Competitive priorities

Decision areas Structural

Infrastructural

Cost Quality Delivery

Process Capacity Facilitates

Quality Organisation Manufacturing planning and control

Flexibility

Vertical integration

1969), Process choice (Hill, 2000), equipment and process technologies (Wheelwright and Hayes, 1985), and process technology (Miltenburg, 2005). Furthermore, different authors include different issues to consider. A comparison of the included considerations within some of these production process related decision areas are gathered in Table 3. A reflection is that the issues included within the decision area remains reasonably similar, although a time span of more than 35 years exits between the authors. In none of the descriptions provided above is automation treated to any larger extent or to any depth. Type of technology is determined on an overall level, and distinction is for example made between general purpose and dedicated technology (Hill, 2000). The decisions are mostly of an ‘‘all-or-none’’ character, which also is noted by Sheridan (2002). 3.2. Manufacturing strategy process The manufacturing strategy process describes the formulation and implementation of a manufacturing strategy. This part of the manufacturing strategy area has attracted less attention in the research community than the manufacturing strategy content (Dangayach and Deshmukh, 2001), in spite of the difficulties associated with strategy implementation Table 3 Comparison of issues included within Production process related decision areas Decision area

Included considerations

Author/s

Plant and equipment

Span of process, plant size, plant location, investment decisions, choice of equipment Scale, flexibility, interconnectedness

Skinner (1969; 1978)

Equipment and process technology Process and Process choice Technology strategy

Process technology

Technology, flexibility jobbing, batch, line Type of technology, leading edge of technology or established technologies, develop technology internally or buying in Nature of the production process, type of equipment, amount of automation, linkages between parts of the production process

Wheelwright and Hayes (1985) Hill (2000) Slack et al. (2001)

Miltenburg (2005)

ARTICLE IN PRESS K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

(Maruchek et al., 1990). Strategy formulation can be done mainly from two different views on competition: A resource- or a market-based view (Gagnon, 1999). With the resource-based view, the issue is to make sure that the resources, capabilities and competencies are properly used as competitive weapons. With the latter view, manufacturing is regarded as a perfectly adjustable system following the rules dictated by the market, while the former view suggests that it is preferable to leverage the unique capabilities of the manufacturing function in order to change the rules of competition. The efforts within operations have been dominated by a market-based view of competition. Lately, a move towards a resource-based view has emerged and it is argued that focusing on developing, protecting, and leveraging a company’s operational resources and advantages to change the rules of competition is a preferred approach (Gagnon, 1999). In practice, it is, however, not a question of either applying the resource- or market-based view of competition when formulating manufacturing strategies. It is essential that requirements from market are considered, and that available manufacturing capabilities are used as wisely as possible. The formulation process is quite similar, independent of the assumed view of competition. With the resource-based view a possible process to formulate a manufacturing strategy is as follows (Gagnon, 1999): (1) An extensive analysis of the manufacturing capabilities. (2) Selection of core capabilities that can provide competitive advantages. (3) Formulation of strategies. The traditional process for strategy formulation represents the market-based view on competition. With this as a starting-point, Hill (2000) proposed the following structure of actions in five steps: (1) Define corporate objectives. (2) Determine marketing strategies to meet these objectives. (3) Assess how different products qualify in their respective markets and win orders against competitors. (4) Establish the most appropriate process to manufacture these products (process choice). (5) Provide the manufacturing infrastructure to support production.

29

This list is of course very stereotyped and in reality there are iterative loops between all of these steps. Hill (2000) stated that there is no shortcut to success in the process of formulating strategies and designing supportive manufacturing processes. Already when defining the corporate objectives, unless the procedure begins when founding the company, the decisions depend on the existing capabilities of the company, which consequently have to be known to the management. Miltenburg (2005) suggests that the actual strategy implementation is guided by an implementation plan, with each decision area taken into consideration. 4. Automation of manufacturing systems A manufacturing system is a collection of equipment, people, and procedures organized to accomplish the manufacturing operations of a company. Within a manufacturing system, a distinction can be made between facilities and support systems (Groover, 2001). The facilities of the manufacturing system consist of the factory, the equipment in the factory, and the way the equipment is organized. The support systems are the procedures used by the company to manage production and to solve the technical and logistics problems encountered in ordering materials, moving work through the factory, and ensuring that products meet quality standards. Automation is the application of mechanical, electronic, and computer-based systems to operate and to control manufacturing. Automation implies that human labour, both cognitively and physically, is replaced by electronic or mechanical devices (e.g. Groover, 2001; Sheridan, 2002). In automated manufacturing systems, the operations are performed with a reduced degree of human participation. The level of automation is often described in discrete steps, i.e. manual, semi-automated, or automated, depending on the task allocation between operators and equipment. 4.1. Automated manufacturing systems Process equipment may be automated in different aspects and automation has a long industrial tradition. Integrated and automated manufacturing systems, such as flexible manufacturing systems (FMS), consisting of a number of integrated machines and control systems were first introduced in the early 1960s (Merchant, 1961). Increasingly

ARTICLE IN PRESS 30

K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

flexible automation of process equipment was enabled as computer technologies advanced rapidly during the 1970s. The machines are equipped with computerized numerical control (CNC) and tool storage systems, which enables short setup times and thus small series production. The manufacturing systems area is constantly developing, and has dramatically improved. Automation of manufacturing support system aims at reducing the amount of manual and clerical effort in product design, manufacturing planning and control, and the business functions of the firm. Manufacturing support systems are technologies that enhance the performance of the processes and sometimes they are absolutely essential in order to achieve the full potential of the manufacturing system. AMT is a collective name for modern and integrated manufacturing technology, such as computer-aided design (CAD), computer-aided manufacturing (CAM), and FMS (Chen and Small, 1996). Besides these technologies, AMT involves a large number of other technologies. The literature on AMT is often related to the physical equipment and its specific characteristics. AMT can be divided according to the different areas of application. A distinction can be made between design applications, manufacturing applications, and administrative applications (Boyer and Pagell, 2000). The different applications involve various types of automated tasks. CAD and computer-aided engineering (CAE) are examples of design applications. Among the manufacturing applications, CAM, robotics, real-time process control systems, FMS, and automated material handling system can be mentioned. Electronic mails, knowledge management system, decision support systems, material requirements planning (MRP), and enterprise resource planning (ERP) are examples of administrative applications of AMT. A common feature of these different applications of AMT is that they more or less involve computers and information technology, replacing human labour. As previously mentioned, one of the major problems with the technology oriented literature is that it concentrates on the specific applications and the potential improvements but fails to explain how to select technological investments that support a business (Hill, 2000). Another problem is that the issue of function allocation, i.e. the possibility of balancing the use of humans and technology is neglected.

4.2. Allocation of physical and cognitive tasks Within human factors engineering, research focus has mainly been on allocation of functions or tasks between humans and technology, considering issues on what to automate with aspects of appropriateness for humans being dominant. Similarly as for the technology-oriented literature a limitation within the human factors engineering domain is that it focuses on the specific applications and does not consider whether different solutions support a business. Fundamental research in task allocation was carried out after the Second World War as a result of war-time experiences of both human and technical limitations. A classical model for function allocation is the Men Are Better At–Machines Are Better At (MABA–MABA) list presented in 1951 (Fitts, 1951) (see Table 4). The original intention of this list was to increase awareness of automation effects. However, in practice, the list has mainly been used for rationalization of mechanization decisions into binary decisions, i.e. to determine whether a task should be automated or not. No other allocation model has replaced the Fitts list in terms of simplicity and understandability (Sheridan, 2002). However, as computers become smarter, the distinction tends to be somewhat less evident. General strategies for function allocation have been suggested in the human factors engineering domain by e.g. Rouse (1991) who describes three automation strategies: comparative, leftover, and economic allocation. Comparative allocation is well described by the MABA–MABA list. Leftover strategies for allocation of tasks relate to the situation where no technical solution can be provided to resolve a specific functionality. Finally, Table 4 The MABA–MABA list (Fitts, 1951) Men are better at

Machines are better at

Detecting small amounts of visual, auditory, or chemical energy Perceiving patterns of light or sound Improvising and using flexible procedures Storing information for long periods of time and recalling appropriate parts Reasoning inductively Exercising judgment

Responding quickly to control signals Applying great force smoothly and precisely Storing information briefly, erasing it completely Reasoning deductively

ARTICLE IN PRESS K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

in economic allocation, if the cost of technology for automating a function is higher than hiring an operator, the function will not be automated even if there is a technical solution. These strategies are static, describing who does what. Once a function is allocated, the responsibility remains either at the machine or with the human (Inagaki, 2003). More recent strategies for designing semi-automated systems emphasize sharing and trading of control, but also scenario-based, dynamic, and adaptive approaches (Sheridan, 2002; Inagaki, 2003; Fallon, 2001). Sharing of control implies that human and machines can function simultaneously whereas trading of control means that the control is passed back and forth. Sharing and trading of control functionalities can be seen in for instance flightdeck and air-traffic control contexts. Three different types of sharing of control can be distinguished: extension, relief, and partitioning (Inagaki, 2003). Extension refers to when for example the computer can extend the capability of the humans, or when the human extend the capabilities of the machines. Allocation based on relief relates to situations where the computer helps the human so that the burden is reduced. The third type of sharing is partitioning, where a function is divided into portions so that machines and humans can deal with mutually complementary parts (Inagaki, 2003). To avoid static situations, dynamically changing allocation of functions is strongly emerging within the automotive industry. We, as automobile ‘‘pilots’’ in new cars, are being flooded with information, warnings, and control options, ranging from automatic braking systems to navigational aids. The strategy is to provide the human with abundant opportunities to dynamically allocate and reallocate tasks. The degree of automation can be described as the level of automation (Sheridan, 1997). Several different scales are available in the literature (e.g. Sheridan, 1997, 2002; Parasuraman et al., 2000) and one example of a 101-scale is presented in Inagaki (2003), see Table 5. The level of automation increases from 1 to 10, where a low level implies mainly manual tasks whereas a high level implies limited or no manual tasks. Application of dynamically changing levels of automation is recognized as an opportunity within the manufacturing industry as well. In the manufacturing context a distinction of automation into computerization and mechanization is found useful (Frohm et al., 2005), relating to the different parts

31

Table 5 Scale of different levels of automation (Inagaki, 2003) Level

Explanation

1

The computer offers no assistance; humans must do it all. The computer offers a complete set of action alternatives, and Narrows the selection down to a few, or Suggests one, and Executes that suggestions of humans approve, or Allows humans a restricted time to veto before automatic execution, or Executes automatically, then necessarily informs humans, or Informs them after the execution only if they ask, or Informs them after execution if it, the computer, decides to. The computer decides everything and acts autonomously, ignoring humans.

2 3 4 5 6 7 8 9 10

of a manufacturing system, i.e. the facilities and the support systems (Groover, 2001). Apart from different levels it is also relevant to consider the different functions that can be automated. Four classes of functions are found describing the human–computer interaction (Parasuraman et al., 2000): information acquisition, information analysis, decision selection, and action implementation. For each of these functions different levels of automation might be appropriate. The possibility of automation at various levels and in different stages is most relevant today with the ever-increasing requirements of adaptability in manufacturing systems (e.g. Parasuraman et al., 2000; Frohm et al., 2005). 4.3. Implementation of automated manufacturing systems Similarly, as for manufacturing strategy, the process of implementation needs consideration. The automation process, i.e. how to implement automation, is described (e.g. Kapp, 1997; Groover, 2001, Baines, 2004), as are issues to consider during implementation (e.g. Gupta and Cawthon, 1996). The USA principle, an automation strategy focusing on aspects to consider prior and during implementation of automated processes is suggested by Kapp (1997). USA stands for Understand, Simplify, and Automate, and the strategy is applicable in different types of automation projects. Another automation strategy is described as

ARTICLE IN PRESS 32

K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

checklists of the possibilities for improving a manufacturing system through automation or simplification (Groover, 2001). This strategy involves issues such as the combination of operations, simultaneous operations, or integration of operations. Groover (2001) furthermore describes an automation migration strategy that is especially useful during new product introduction. The automation migration strategy is a plan for evolving the manufacturing systems used to produce new products as the demand grows. The migration strategy involves three phases: manual production, automation of single stations, and finally automated integrated production. Following the migration strategy provides several advantages, although no guidance is provided concerning the appropriate task allocation and level of automation. 5. Applied practical automation approaches and their consequences A number of case studies were performed in order to study how they worked with automation issues linked to automation decisions. They were all in the process of investing in new manufacturing capacity, but the degree of maturity in strategy process and content showed great differences. We can see that two cases were quite successful whilst the others did not show the intended outcome. 5.1. Case A One example when automation was the manufacturing strategy was described by a manufacturing company in the telecom industry, company A (Johansen et al., 2001; Winroth, 2004). Decisions were made on automation of the final assembly line for cellular phones. The order from top management was to automate as much as possible. In some cases, the industrial engineers in charge of the investment project did not have to make financial justifications of the investments. The company, however, lacked competence and experiences from automated assembly systems and the equipment supplier had a large number of engineers for installing and starting-up of the assembly line. The problems associated with the new automated assembly line were neglected by the management, which contributed to the poor outcome and the following drastic cost reductions.

5.2. Case B Another example is provided from a truck manufacturer, company B, nowadays well known for its good performance in manufacturing. One example, however, where the automation strategy was not linked to the company’s manufacturing capabilities, was when the company in one of its plants introduced industrial robots and new sophisticated equipment in order to improve flexibility, increase productivity, and reduce manufacturing cost. The equipment was, however, too complicated and problems occurred (Sa¨fsten, 1998). After a few years, the company removed the robots and went back to more manual tasks, which led to an increase in productivity by 50% (Magnerot, 2002). 5.3. Case C A manufacturer in the wood-manufacturing industry, company C, made a comparatively large investment in a new production line. The idea emanated from the production manager, but the managing director was also directly involved in the project. The production line is designed in two sections, the first half is fully automated and the second is entirely manual. As the company’s policy is to be very flexible, and meet the customers’ demands on more or less customized products, enormous problems occur in balancing the manual part of the line since the work content varies. Huge buffers, which increase the cost for work in process, are built up and have to be taken care of, thus causing extra cost. In this case, the investment was not correlated to the long-term business and manufacturing strategies. The lack of competence in handling automated equipment is also a problem. 5.4. Case D A successful approach, when the automation decisions are made in congruence with the other decisions in a manufacturing strategy, is described by a manufacturer of springs, company D. Their product areas are industrial springs, strip springs, automotive springs, and gas springs. The present plant is fairly new since the old one was destroyed by fire in year 1996. The rebuilding of the plant is described in Bellgran and Sa¨fsten (2005). The main key to success, in the company’s opinion, has been a combination of good industrial engineering and

ARTICLE IN PRESS K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

business development. Important success factors are considered to be in-house product development and good control of the manufacturing process, which includes tool manufacturing and methods planning. Their competitive priorities are short development time, flexibility, and reliable deliveries. The start of the new factory was in fact an opportunity to create something really well planned. The analysis work in connection with the planning of the new plant included areas such as:

     

Customer segmentation, qualifiers and winners. Product mix, processing position. Technical resources and their main characteristics. Product-position/-profile. Production flow analysis. Decision management.

The product range was categorized and a product profiling was performed, where products and markets, manufacturing, and different possible process choices were matched against each other. The result was that the plant was organized in four different production flows, which are well suited for each category of products. The delivery precision and reliability have improved considerably compared with the old plant. Other consequences are reduced lead-time, from 4 to 5 weeks to 10 days, the productivity has more than tripled, the capacity doubled, and the production area has been reduced to half. This case is an example where the automation decisions were linked to the manufacturing strategies, which led to a successful outcome. 5.5. Case E Another success is drawn from the office furniture industry. In year 1995, the company experienced a crisis since their larger competitors showed much higher productivity than this small (22 employees) independent manufacturer. The manufacturing showed a very low level of automation and about 90% of the manufacturing cost emanated from cost of labour and only 10% from depreciation of investments. Most equipment was general purpose, which demanded a high manning level. The premium segment of office furniture demands high quality and a high degree of customized design. The manufacturing equipment was quite old and could not cope with the output demands and it was not

33

possible to meet the customization requirements. The company decided to invest in state-of-the-art equipment. They have seen to training the operators in the new machines, even if the operators sometimes were quite old and had a rather short time to retirement. Today the level of education is comparatively high and all employees have at least 2–3 years of professional education and some have university degrees. The operators do their own CAM conversion of the drawing data at the office and data is transferred on-line to the machines. Their main competitive priorities are:

    

Design adaptation. Flexibility and customization. Cost. Efficiency and productivity. Lead times and delivery precision.

The company does not have its own maintenance resources, but they manage through close collaboration with external parties, such as one major supplier of equipment, who also is collocated with the factory. The company has no ambition to invest in full automation and when they invest in a new machine they also see to hiring one additional operator. The main reason for this standpoint is that they consider it crucial to use the human eye for colour and pattern matching. 6. Automation strategies—part of the manufacturing strategy Two different perspectives on automation strategies have previously been identified (Winroth et al., 2007). The first perspective is when the overall manufacturing strategy is equal to an automation strategy, i.e. the strategy is automation. With this perspective automation is a functional strategy on its own, in parallel with for example with the market strategy and the R&D strategy, see Fig. 1(A). The other perspective, which has proven to be the most successful (see e.g. cases D and E), is when decisions concerning automation are treated as one of several decisions in a manufacturing strategy, see Fig. 1(B). This is the perspective mainly communicated in the operations management literature. Among other decisions during the manufacturing strategy formulation, one question is to what degree different tasks should be automated (Slack et al., 2001). When automation is one of several aspects considered in the manufacturing strategy, the

ARTICLE IN PRESS 34

K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

Fig. 1. Two perspectives on automation and manufacturing strategy.

decisions concerning automation is a consequence of the manufacturing capabilities (such as cost, quality etc.) that the company wants to achieve. Furthermore, the automation decisions are part of the decision area concerning the production process. Automation is traditionally treated rather superficially within the area of manufacturing strategy (e.g. Miltenburg, 2005, Hill, 2000). To fully utilize the manufacturing potentials provided by automation, improvement and refinement of the automation decision on a strategic level is required. Decisions concerning automation require consideration of the possible advantages of choosing between different levels of automation, with guidance from the specific situation. Within human factors engineering, knowledge is available concerning aspects to consider when selecting the level of automation based on how to allocate functions between human systems and technical systems (e.g. Parasuraman et al., 2000; Sheridan, 2002; Fallon, 2001). This knowledge can support and complement the manufacturing strategy theory. A tentative framework, indicating the central role of function allocation and different levels of automation in the manufacturing strategy content and process is suggested, see Fig. 2. Refinement of manufacturing strategy content and process is achieved with support from human factors engineering. Within human factors engineering the issue of task allocation, or function allocation, is addressed (e.g. Sheridan, 2002; Fallon, 2001). Different levels of automation, appropriate for different functions and in different situations, are elaborated upon. Among other things, scales of different levels of automation are provided (e.g. Sheridan, 2002; Parasuraman et al., 2000) giving a nuanced picture of automation. The application of these experiences, and this

Fig. 2. Refinement of manufacturing strategy content and process with support from human factors engineering.

knowledge, when determining the content of the decision areas and the strategy process, supports the competitive priorities and thereby the possibilities of competitiveness for manufacturing companies. Automation decisions have to support a business and the overall goals in a company in order to deliver expected advantages (e.g. Hill, 2000; Boyer et al., 1996); the ‘‘none-or-all fallacy’’ needs rejection. Furthermore, some of our empirical studies, cases A–C, indicate that when decisions concerning automation are made without considering the context, i.e. the other relevant decision areas, the long-term result is not satisfactory. According to Hayes and Wheelwright (1984), process choice, including choice of technology level and automation, is one of the eight strategic decision areas that are important for the success of a manufacturing organization. All these decision areas are, however, closely interlinked consequently leading to tradeoffs, i.e. it is impossible to achieve the highest performance in all areas at the same time (e.g. Skinner, 1969). Since the areas are interlinked, the choice of a certain level of automation calls for a certain skill level of personnel, the component

ARTICLE IN PRESS K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

supply needs to be carried out in a certain manner, and quality management needs to be considered, etc. Decisions concerning automation affect all of the decision areas, and, therefore, have to be considered in conjunction. Decisions about automation directly impact several other decision areas:











The vertical integration is very important since a problem at the supplier will directly lead to problems at the systems integrator and with an automated system these problems will have to be taken care of probably without human support. The quality management system should be supportive to the technology level that we choose, including self-adjusting Statistical Process Control, SPC, and adaptive control. The skill level of the personnel needs to be in congruence with the technology level for managing the system, doing programming tasks etc. Some of the work tasks involved in a highly automated manufacturing system may be simple routine tasks, but it is also likely that new and very advanced tasks are created. With more competent personnel who have delegated responsibility, the corresponding authority needs to be included. If the organizational structure is highly hierarchical, the full potential of the personnel will be lost. The system for production planning and control also needs to be linked to the level of automation.

The process choice is often very much dependent on the actual level of automation in order to create stable processes that are able of providing the desired output. The design of the facilities and the layout are closely linked to the level of automation as well as how to handle the sourcing issues (Groover, 2001). Groover (2001) furthermore points out that level of automation influences the longterm strategies of the company related to the level of competence and where to locate production. It also influences several output factors such as quality, delivery issues, and flexibility. The choice of automation level needs consideration already when starting up the design work. When the automation strategy is part of the manufacturing strategy, decisions about automation are one of several decisions in the manufacturing strategy. The driving forces behind the automation decisions can either be the identified

35

needs and requirements from the market that best can be realized with some level of automation and/ or an identified potential within the company to compete with a certain level of automation. The processes are likely to follow the traditional strategy formulation processes described by Hill (2000), which represents a market-based view on competition, or the formulation process based on the resource-based view of competition (Gagnon, 1999). When automation at any degree is involved, the issue of automation implementation needs consideration of the automation strategies suggested by for example Groover (2001) and Kapp (1997). Irrespective of whether a market-based view or a resource-based view on competition is applied, the implementation plan, suggested by Miltenburg (2005), can be supported with considerations related to the different automation strategies. It is, however, important to remember that the USA principle (understand, simplify, automate) also involves tasks preferably done prior to the decision on appropriate automation level. 7. Conclusion Two different perspectives on automation strategies have previously been identified (Winroth et al., 2007). The most successful results are achieved when decisions concerning automation are treated as one of several decisions in a manufacturing strategy. The manufacturing strategy content and process, however, was found to need refinement in order to fully utilize the potential of automation. This paper has elaborated on the possibilities of refinement of manufacturing strategy content and process with support from the human factors engineering domain. The overall ambition is to improve the support from automation on manufacturing competitiveness, an issue that needs linkage between operational issues and strategic decision. A tentative framework is suggested (see Fig. 2) illustrating the suggested role of function allocation during the strategy formulation process, i.e. a link between operational concerns and strategic decisions. Experiences and knowledge from human factors engineering domain is considered during strategy formulation. The result is a more nuanced picture of automation. The consequences from different levels of automation are considered for all the different decision areas. This enhances the possibility that decisions concerning automation are appropriate for the situation and

ARTICLE IN PRESS 36

K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

thereby hopefully supportive for the overall performance of the company. The issue of appropriate automation is important since fit between manufacturing technology and business strategy is found to affect performance of manufacturing firms (e.g. Kotha and Swamidass, 2000). Thus, the effects from considering automation strategy as part of the manufacturing strategy, nuanced automation decisions through operational and strategic linkages, and automation concerns in all of the decision areas, is potentially supporting improved manufacturing performance and competitiveness. With appropriate levels of automation, in every specific situation, most positive effects on the manufacturing performance is expected, see Fig. 3. If the level of automation is too low, i.e. underautomation, or too high, i.e. over-automation, the potential of automation is not fully utilized for the

support of manufacturing competitiveness. An appropriate level of automation, ‘rigthomation’, contributes positively in several respects, whereas the effects from both under and over automation can have negative effects on manufacturing performance. The effects given on the left hand in Fig. 3 are mainly based on the empirical material presented in this paper, and are examples of possible effects the chosen level of automation can have on the manufacturing performance and competitiveness. Further research is, however, required to test the applicability of the suggested frameworks (Figs. 2 and 3) and the actual improvements that are possible to achieve. The manufacturing strategy formulation process also needs further support in terms of guidelines indicating under which circumstances certain levels of automation are appropriate in specific manufacturing situations.

Fig. 3. Appropriate level of automation, ‘rigthomation’, is expected to have positive effects on the manufacturing performance.

ARTICLE IN PRESS K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

References Baines, T., 2004. An integrated process for forming manufacturing technology acquisition decisions. International Journal of Operations & Production Management 24 (5), 447–467. Bellgran, M., Sa¨fsten, K., 2005. Produktionsutveckling–Utveckling och drift av produktionssystem. Studentlitteratur, Lund, Sweden. Boyer, K.K., Pagell, M., 2000. Measurement issues in empirical research: Improving measures of operations strategy and advanced manufacturing technology. Journal of Operations Management 18, 361–364. Boyer, K.K., Ward, P.T., Leong, G.K., 1996. Approaches to the factory of the future–An empirical taxonomy. Journal of Operations Management 14, 297–313. Chang, T.-C., Wysk, R.A., Wang, H.-P., 2005. Computer-Aided Manufacturing, third ed. Prentice-Hall, Upper Saddle River. Chen, I.J., Small, M.H., 1996. Planning for advanced manufacturing technology–A research framework. International Journal of Operations & Production Management 16 (5), 4–24. Dangayach, G.S., Deshmukh, S.G., 2001. Manufacturing strategy: Literature review and some issues. International Journal of Operations & Production Management 21 (7), 884–932. Dean, J., Yoon, W., Susman, G., 1992. Advanced manufacturing technology and organization structure: Empowerment or subordination? Organization Science 3, 203–229. Fallon, E.F., 2001. Allocation of Functions: Past, present and future perspective. In: Karwowski, W. (Ed.), Encyclopaedia of Ergonomics and Human Factors. Taylor & Francis, Florence, USA. Fitts, R.M., 1951. Human Engineering for an Effective Airnavigation and Traffic-control System. National Research Council, London. Frohm, J., Lindstro¨m, V., Bellgran, M., 2005. A model for parallel levels of automation within manufacturing. Proceedings of the 18th International Conference on Production Research (ICPR), Fisciano, Italy, July 31–August 4. Fujimori, T., 1990. Development of the flexible assembly system SMART, Proceedings of the 21st International Symposium on Industrial Robots (ISIR), Copenhagen, Denmark, October 23–25. Gagnon, S., 1999. Resource-based competition and the new operations strategy. International Journal of Operations & Production Management 19 (2), 125–138. Granell, V., Frohm, J., Winroth, M., 2006. Controlling Levels of Automation–A model for Identifying Manufacturing Parameters, Proceedings of the Ninth IFAC International Federation of Automatic Control Symposium on Automated Systems Based on Human Skill and Knowledge, Nancy, France, May 22–24. Groover, M.P., 2001. Automation, Production Systems, and Computer-Integrated Manufacturing, second ed. PrenticeHall, Englewood Cliffs, NJ, USA. Gupta, M., Cawthon, G., 1996. Managerial implications of flexible manufacturing for small/medium sized enterprises. Technovation 16 (2), 77–83. Hayes, R.H., Clark, K.B., 1995. Why some factories are more productive than others. In: Pisano, G.P., Hayes, R.H., (Eds.), Manufacturing Renaissance. Harvard Business School Press, Boston, MA, pp. 133–149.

37

Hayes, R.H., Pisano, G.P., 1994. Beyond World-Class: The New Manufacturing Strategy. Harvard Business Review January– February, 77–86. Hayes, R.H., Wheelwright, S.C., 1984. Restoring Our Competitive Edge: Competing Through Manufacturing. Wiley, New York. Hill, T., 2000. Manufacturing Strategies, third ed. Irwin McGraw-Hill, Boston. Inagaki, T., 2003. Adaptive Automation: Sharing and Trading of Control. In: Hollnagel, E. (Ed.), Handbook of Cognitive Task Design. Lawrence Erlbaum Associates, Inc., Publishers, London. Johansen, K., Winroth, M., Bjo¨rkman, M., 2001. An economical analysis on a strategic investment in an assembly line: Case study at Ericsson Mobile Communications AB, Linko¨ping, Sweden. Proceedings of the 16th International Conference on Production Research (ICPR), Prague, Czech Republic. Kapp, K.M., 1997. The USA Principle, APICS—The Performance Advantage June, pp. 62–66. Kotha, S., Swamidass, P.M., 2000. Strategy, advanced manufacturing technology and performance: empirical evidence from US manufacturing firms. Journal of Operations Management 18, 257–277. Magnerot, D., 2002. Sko¨nt att ma¨nniskor beho¨vs, Ny Teknik June 5th Swedish Weekly Technical Paper. Makino, H., Yamafuji, K., 1988. Trends in automatic assembly in Japan. Proceedings of the Ninth International Conference on Assembly Automation, London, UK. Maruchek, A., Pannesi, R., Anderson, C., 1990. An exploratory study of the manufacturing strategy process in practice. Journal of Operations Management 9 (1), 101–123. Mehrabi, M.G., Ulsoy, A.G., Koren, Y., 2000. Reconfigurable manufacturing systems: Key to future manufacturing. Journal of Intelligent Manufacturing 11, 403–419. Merchant, M.E., 1961. The manufacturing system concept in production engineering research. Annals of the CIRP 10 (2), 77–83. Miles, M.B., Huberman, A.M., 1994. Qualitative Data Analysis: An Expanded Sourcebook. Sage Publications, Inc., Thousand Oaks, California. Miltenburg, J., 2005. How to Formulate and Implement a Winning Plan. Productivity Press, New York. Parasuraman, R., Sheridan, T.B., Wickens, C.D., 2000. A model for types and levels of human interaction with automation. IEEE Transactions on Systems, Man, and Cybernetics Part A: Systems and Humans 30 (3), 286–297. Roth, A.V., Miller, J.G., 1992. Success factors in manufacturing. Business Horizons 35 (4), 73–81. Rouse, W.B., 1991. Design for Success—A Human Centred Approach to Designing Successful Products and Systems. Wiley, New York. Sa¨fsten, K., 1998. Requirements and Strategic Preconditions for Efficient Assembly, Linko¨ping Studies in Science and technology, Thesis No. 707, Linko¨ping, Sweden. Sa¨fsten, K., 2002. Evaluation of assembly systems—An exploratory study of evaluation situations. Ph.D. Thesis, Linko¨ping Studies in Science and Technology, Dissertation No. 756, Linko¨ping, Sweden. Sheridan, T.B., 1997. Task analysis, task allocation and supervisory control. In: Helander, M., Landauer, T.K., Prabhu, P. (Eds.), Handbook of Human–Computer Interaction, second ed. Elsevier, Amsterdam. Sheridan, T.B., 2002. Humans and Automation: System Design and Research Issues. Wiley, Santa Monica.

ARTICLE IN PRESS 38

K. Sa¨fsten et al. / Int. J. Production Economics 110 (2007) 25–38

Skinner, W., 1969. Manufacturing—missing link in corporate strategy. Harvard Business Review May–June, 136–145. Skinner, W., 1978. Manufacturing in the Corporate Strategy. Wiley, New York. Slack, N., Chambers, S., Johnston, R., 2001. Operations Management, third ed. Pearson Education, Prentice-Hall, England. Swink, M., Way, M.H., 1995. Manufacturing strategy: propositions, current research, renewed directions. International Journal of Operations & Production Management 15 (1), 4–26. Ward, P.T., Bickford, D.J., Leong, G.K., 1996. Configurations of manufacturing strategy, business strategy, environment and structure. Journal of Management 22 (4), 597–626. Wheelwright, S.C., Hayes, R.H., 1985. Competing through manufacturing. Harvard Business Review, January–February 4, 115–125.

Wheelwright, S.C., 1984. Manufacturing Strategy: Defining the Missing Link. Strategic Management Journal 5, 77–91. Winroth, M., 2004. On manufacturing strategies—Competing through inter-organizational collaboration. Ph.D. Thesis, Linko¨ping Studies in Science and Technology, Dissertation No. 860, Linko¨ping, Sweden. Winroth, M., Sa¨fsten, K., Stahre, J., 2007. Automation strategies – Existing theory or ad hoc solutions? International Journal of Manufacturing and Technology Management 11 (1), 98–114. Yin, R.K., 1994. Case Study Research Design and Methods. Sage Publications Inc., Thousand Oaks, California. Yu, J., Yin, Y., Sheng, X., Chen, Z., 2003. Modelling strategies for reconfigurable assembly systems. Assembly Automation 23 (3), 266–272.