JUST-IN-TIME, LEAN PRODUCTION, AND COMPLEMENTARY PARADIGMS:LEAN PRODUCTION AND EXTENSIONS OF JIT

LEAN PRODUCTION AND EXTENSIONS OF JIT

Lean Production

The term lean production was introduced by Krafcik (1988) and the famous book, The Machine That Changed the World (Womack et al. 1990). These publications present the results of a major MIT study to identify systematically best practices of Japanese and other automobile manufacturers world- wide. Lean production is ‘‘lean’’ in that it uses half of the various production resources (labor, manufacturing space, tool investment, engineering hours, inventory, etc.) used in the Ford-style mass production that was prevalent into the 1980s.

The essence of lean production may be summarized into four aspects: (1) lean plant; (2) lean supplier network; (3) lean product development; and (4) relationship with distributors and customers. Of these, the lean plant forms the original core and may be considered as equivalent to the JIT / TPS concept presented earlier. The remaining three aspects delineate how the lean plant interacts with and mutually influences other entities. Thus, lean production may also be considered as an extended paradigm of JIT that includes new intraorganizational and interorganizational aspects. It is interesting to note that Toyota itself was influenced by the systematic, encompassing framework of lean pro- duction as presented by the MIT study. Though Toyota’s practices indeed formed the basis for the lean production model, many of those practices had evolved gradually in response to local organi- zational needs or involved historically separate organizational units. The lean production model con- sequently provided a conceptual thread for understanding its many aspects as a synergistic, integrated whole.

Lean supplier network, referring to various innovative practices in supplier relationships, can be considered as an interorganizational extension of JIT. It is well known that Toyota reorganized its suppliers into functional tiers in the 1950s. Its aim was to provide suppliers with incentive for improvement and promote learning mechanisms among suppliers and workers alike while maintaining long-term relationships. In this system, different responsibilities are assigned to the suppliers in each tier. First-tier suppliers are responsible not only for the manufacture of components, but also for their development while interacting and sharing information with Toyota and other first-tier suppliers. In turn, each first-tier supplier forms its own second tier of suppliers. They are assigned the job of fabricating individual parts and occasionally assist in their design, within the confines of rough specifications provided by the first-tier companies. In many cases, even a third tier of suppliers may exist to manufacture parts according to given specifications.

Roughly only 30% of Toyota’s parts production is done in-house, with the remainder allotted to the supplier network. This high degree of outsourcing affords Toyota flexibility in coping with change. At the same time, it underscores the criticality of a competitive supply chain. Toyota’s supplier network is neither a rigid, vertical integration nor an arm’s-length relationship. Furthermore, it at- tempts to maintain a delicate balance between cooperative information sharing and the principle of competition. For this purpose, Toyota has employed several organizational strategies.

One approach for promoting information sharing is the formation of supplier associations and the cross-holding of shares with supplier-group firms and between first-tier suppliers. Another is the two- way sharing of personnel with Toyota and between supplier-group firms.

As an approach for encouraging competition, there is ample opportunity for a supplier to move up the ladder if it exhibits very high performance in quality, cost, delivery, and engineering capability. A parts-manufacturing specialist, for example, can expand its role to include the design of parts and even integrated components. Further incentive is provided by a degree of competition between sup- pliers over their share of business for a given part. Contrary to common assumption, parts are not sole-sourced except for certain complex systems that require considerable investments in tools. Rather, the business is typically divided between a lead supplier and one or more secondary suppliers. Each supplier’s relative share of business may grow or shrink depending on performance. By providing substantial incentives to suppliers, this flexible system has spurred a variety of improvements relating to quality, cost, and JIT delivery of parts.

Another aspect of the lean production enterprise concerns the way that new products are designed and developed. Lean product development differs from traditional approaches on four key dimensions: leadership, teamwork, communication, and simultaneous development (Womack et al. 1990). In the case of automobile development, hundreds of engineers and other staff are involved. Consequently, strong leadership is the first condition for doing a better job with less effort. Toyota adopted the shusa system, also known as the heavyweight project manager system (Clark and Fujimoto 1991). The shusa (literally the chief engineer) is the leader of a project team whose job is to plan, design, and engineer a new product as well as to ramp up production and deliver it to market. Great re- sponsibility and power are delegated to the shusa. Most importantly, the shusa assembles a flat team with members coming from the various functional departments throughout the company, from mar- keting to detail engineering. Though retaining ties to their original department, the team members are now responsible on a day-to-day basis to the shusa.

The development process used to explore the optimum design solution has been called by names such as set-based engineering (Liker et al. 1995). It can be defined as an approach for engineers to explicitly consider and communicate sets of design alternatives at both conceptual and parametric levels within clearly defined constraints. They gradually narrow these design sets by eliminating inferior alternatives until they eventually freeze the detailed product specifications based on feasibility studies, reviews, and analysis. This broad, sweeping approach helps the design team avoid premature design specifications that might appear attractive based on narrow considerations but that would suboptimize the overall design. Furthermore, set-based engineering facilitates sharing of the product’s overall image among all members involved in the project, making it possible to develop different aspects of the product simultaneously.

This simultaneous development approach is now generally known as concurrent or simultaneous engineering. It is defined as the overlapping or parallel development of what otherwise would be sequential jobs. For example, the design of a production die ordinarily is not initiated until the detail engineering is completed for the part to be produced. In concurrent engineering, however, the die design is commenced even before the part design in finalized by utilizing shared knowledge con- cerning the approximate dimensions of the part, its requirements, and its design process. This over- lapping design process is made possible by close communication between the sequential tasks and is somewhat analogous to the exploitation of off-line or external setup in order to reduce production lead time. Concurrent engineering is also employed with outsourced components by sharing personnel and information between Toyota and its suppliers.

Through concurrent engineering and other approaches, lean product development has achieved astonishing results, requiring only one half or less of the time and human resources traditionally needed to develop a new car. Accordingly, it has become a cornerstone of time-based competition. Further means for improvement lie in the full exploitation of information technologies. In particular,

CAD, CAE, and electronic interchange of technical data facilitate the sharing of information and have further potential for supporting concurrent engineering, particularly in later development stages.

The final aspect of lean production is the relationship with distributors and customers. As it did with its supplier network, Toyota built a distribution network that incorporated the dealers into the production system. In fact, the dealer is considered the first step in the kanban system. It sends orders for presold cars to the factory for delivery to specific customers about two weeks later, thereby initiating the pull signal for production. For this to be workable, the dealer works closely with the factory to sequence orders in a way that the factory can accommodate, just as the factory works closely with its suppliers. At the same time, dealers have played another role in assessing customer needs and relaying that information to Toyota for consideration in designing new models. Dealers have done this by making house calls directly to customers and increasingly by collecting customer information in databases maintained at the dealers.

Through these innovative practices, Toyota’s lean production system formed a complete supply chain from source to user. Its system integrated suppliers and distributors with manufacturing and involved them in product development. In so doing, the lean production model has provided a pro- totype of the business paradigms of concurrent engineering and supply chain management. For an examination of the implementation of lean production in North American companies, see Liker (1998) and Liker et al. (1999).

Theory of Constraints (TOC) and JIT

The concept and practice of kaizen are features common to TQM, TPM, JIT, and lean production. However, despite great effort expended for continuous improvement throughout their organizations, many Japanese companies in the 1990s faced difficulty in improving their financial bottom line, including even the leaders in JIT practice. This was largely due to Japan’s prolonged economic recession, but many companies would have benefited from a more focused approach to their im- provement efforts.

One such approach lies in the Theory of Constraints (TOC). Devised by Eli Goldratt (e.g., Goldratt 1990; Goldratt and Cox 1992), TOC can be viewed as an inclusive philosophy and methodology of improvement, which includes bottleneck-focused scheduling methods, performance measurement sys- tem, thinking process, and problem-solving tools. Its core idea is that every system, such as a for- profit company, must have at least one constraint. Since a constraint is a factor that limits the system from getting more of whatever it aims to achieve, a company that wants more profits must manage its constraints. See Spencer and Cox (1995) for an analysis of the history and scope of TOC and its relation to optimized production technology (OPT), as well as Umble and Srikanth (1990), Noreen et al. (1995), and Dettmer (1997) for detailed discussions of TOC’s sub-areas of scheduling / logistics, performance measurement, and thinking process, respectively. A review of TOC-related journal publications is given by Rahman (1998).

Among TOC’s many techniques and tools, the five-step focusing process (e.g., Goldratt and Cox 1992) is the one most closely related to JIT:

Step 1. Identify the system constraint(s).

Step 2. Decide how to exploit the system constraint(s), i.e., better utilize the constraint’s existing capacity.

Step 3. Subordinate everything else to the above decision, i.e., align the rest of the system to work in support of the constraint.

Step 4. Elevate the constraint(s), i.e., improve / increase the constraint’s capacity.

Step 5. If a constraint has been broken, go back to step 1. Do not allow inertia to cause a system constraint.

Of these, steps 1, 4, and 5 are practically equivalent to the continuous-improvement logic of JIT if constraint is considered in the narrow meaning of a production bottleneck. One could even consider these five focusing steps as JIT concepts in combination with steps 3 and 4’s unique bottleneck scheduling logic.

Despite this similarity, two important differences should be noted. The first difference relates to the meaning of the term constraint and to understanding which goal is being constrained. In JIT, the constraint is tantamount to a production bottleneck, and the primary focus is placed on inventory level. In the case of TOC, the issue is not inventory level per se, but rather achieving the company’s goal of making profit by means of increasing throughput, reducing expenses, and / or reducing inven- tory.

To understand TOC’s approach to improving company profitability, consider the often-drawn analogy between a system and a chain, where the strength of the chain corresponds to the company’s profit. To improve the strength of the chain, one must identify the weakest link and concentrate effort on strengthening it, instead of applying effort uniformly over all the links. This analogy implies that the focus for continuous improvement must be on the area that will bring about the greatest benefit in relation to the effort extended. Furthermore, this area for improvement may lay outside the man- ufacturing plant because the constraint may not be a physical one but may involve managerial policies or market factors. Consequently, in such a situation, plant-wide improvement efforts in ‘‘carpet bomb- ing’’ fashion may not lead to an improved bottom line for the company.

The second point of difference concerns the focus of improvement efforts. Whereas JIT, TQM, and TPM concentrate on changing the production system as the means to improve it, TOC explicitly considers options for increasing profits while making the constraint work more effectively as it is. This concept is embodied in steps 2 and 3 of the five-step focusing process. In particular, step 3 corresponds to the famous scheduling solution called drum-buffer-rope (DBR). As illustrated in Fig- ure 5(a), DBR is easily understood using Goldratt’s analogy of a Boy Scout troop. Each Scout stands for one workstation or process, while the distance the troop moves during a specific period is through- put earned and the length of the Scout troop column corresponds to work-in-process. Under the condition that we cannot eliminate statistical fluctuations and disruptions in each Scout’s hiking performance, the objective is to have the Scout troop advance as much distance as possible while keeping the troop’s length short.

The solution to this problem lies in having the slowest Scout (constraint) set the pace of the troop by beating a drum and tying a rope with some slack (work-in-process or time buffer) between him and the leading Scout. The rope prevents the troop length from increasing unnecessarily. At the same time, the slack in the rope allows the slowest Scout to work at his full capability without any interference from disruptions in the preceding Scouts’ performance, thereby enabling the troop as a whole to advance the maximum attainable distance.

For reference, the corresponding analogy for a kanban system is shown in Figure 5(b). In a kanban system, all adjacent pairs of Scouts are tied together by a rope whose length represents the number of kanbans in circulation, and the pace of the drum beat must be constant, in keeping with JIT’s prerequisite of leveled production. It should be noted that there is no individual Scout who corre- sponds to a constraint or bottleneck. Instead, once any Scout experiences difficulty walking with the given length of rope, that Scout is identified as a constraint and countermeasures are taken straight- away to increase his capability. Next, the length of rope between each pair is further shortened and another constraint is discovered to repeat the cycle of continuous improvement. In this sense, kanban should be considered as a means to expose the constraint, rather than a means to schedule it and fully exploit its existing capability.

In sum, TOC adds a beneficial and indispensable viewpoint to the 3T’s in that it helps to clarify the goal of improvement and to identify where improvement should be focused so as to achieve maximum financial benefits from a global optimum standpoint. In addition, TOC complements JIT, TQM, and TPM by emphasizing that improvement should not always be the first option. Rather, quick returns are often possible from first exploiting the constraint as it is, making it work to its own utmost limit.

Applications to Service Industries

As exemplified by such terms as the production-line approach to service or the service factory, manufacturing and service industries have a history of mutually influencing each other’s theory and practice. Despite their manufacturing industry origins, JIT, TQM, TPM, and TOC also have appli- cations to service industries. The purpose of this section is briefly to identify some of these application issues and provide references for further information.

JIT and lean production concepts such as rationalization of process flows, batch size reduction, demand leveling, and multifunctional workers are quite applicable to many service environments. Duclos et al. (1995) provide a review of JIT practices in services as well as references for articles describing their implementation. For example, Mehra and Inman (1990) describe an overnight pack- age delivery company’s implementation of JIT for its business supplies, and Billesbach and Schnei- demans (1989) examine the application of JIT in administrative areas. A broader discussion of the application of lean production principles to service operations is provided by Bowen and Youngdahl (1998), wherein they describe how service businesses such as Taco Bell, Southwest Airlines, and Shouldice Hospital have mastered ‘‘lean’’ service. Southwest Airlines, for example, is known for its rapid turnaround (setup) of planes between flights—about three times faster than the industry average. Service companies should not wrongly assume that production efficiency and customer responsiveness are tradeoffs, as they were in the mass production paradigm. Rather, lean production principles can be used to eliminate non-value-added activities from service processes, integrate the value chain, and increase flexibility and responsiveness.

Of various other JIT and TQM-related concepts, poka-yoke merits special attention. In many services, due to the simultaneity of a service’s creation and consumption, mistakes are readily ap- parent to the customer and it is not possible to carry out inspection or rework. Examples of mistake- proofing in service environments range from color coding of patients’ medical files to techniques for differentiating drinks with similar colors. Further discussion and examples of poka-yoke in services are provided by Schvaneveldt (1993) and Chase and Stewart (1994).

While TPM is practiced in the administrative areas of many leading manufacturers, it has received scant attention from service industries. This is understandable given the different nature of technology and equipment in most service environments. However, a precursor to TPM called 5S is very appro- priate for services. As briefly described in the TPM section above, 5S is a Japanese acronym for five good housekeeping practices for organizing and cleaning the workplace. These practices involve all employees, not just designated maintenance or custodial staff, and help to establish an environment of self-responsibility and problem awareness among employees.

Finally, several aspects of TOC have application to service industries, particularly the five-step focusing process for system constraint identification and management. In addition to reviewing the emerging literature on TOC in services, Siha (1999) provides a framework for interpreting and ap- plying TOC concepts to different service industries. An interesting case study of TOC in an engi- neering services firm is given in Motwani and Vogelsang (1996), which describes how the firm identified its constraint in the survey department and took measures to increase throughput. In another case (Olson 1998), a security and alarm company was unable to meet market demand. It determined that installation technicians were the bottleneck resource and consequently redesigned the alarm- installation process to allow technicians to work in teams and perform tasks in parallel. TOC has even been applied in an accounting firm (Green and Larrow 1994). After identifying the constraint to be the tax department, the firm found ways for the rest of the organization to support the work of the tax department better without increasing staff or work hours.