The group structure also turned out to be uniquely supportive of Ohno’s concept of target costing, where Toyota Motors at the top of the pyramid determined the value of a given component to the customer and then worked backwards with the supplier to figure out how to remove enough cost to produce the part at the target cost with an acceptable profit. As we will see in a moment, the best way to remove cost was almost always to embrace the Toyota Production System (TPS).
As Toyota group supplier costs fell, the 190 firms soon discovered that they could make much more money selling to customers other than Toyota who did not understand the logic of lean production. Toyota soon began to receive a cross-subsidy from all of its competitors except Nissan, which Toyota’s core suppliers were barred from selling to until 1994.
Completing the Revolution in Production
By the mid-1960s, Ohno had finally pushed his ideas all the way through Toyota’s own production facilities. The logical next step was for all Toyota suppliers to begin delivering parts Just-in-Time. However, as delivery frequency was increased in response to kanban signals, Toyota discovered that its suppliers were relying on finished goods warehouses filled with small piles of parts assembled far in advance for their hourly or several-times-a-day shipments. The piles were created from large production batches because the suppliers had no idea how to produce in small lots to replenish the amounts withdrawn from stocks several times a day by Toyota.
In 1969, Ohno therefore directed a new group of direct reports he had trained personally, the Production Research Office (now called the Operations Management Consulting Division [OMCD]), to set up mutual-help groups among Toyota’s forty-two largest and most important suppliers. The companies were divided into six groups of seven, each with a team leader from one of the companies. The groups were asked to conduct one major improvement activity each month between them, with the technical assistance of OMCD. The results of the activities were then to be examined by senior executives of the other six firms whose task was to offer suggestions on how the activity might be improved even further. Next, the suppliers were asked to establish their own OMCDs and get on with the task of making every activity lean. Toyota pulled the transformation along by demanding continuing reductions in part costs on every part every year from every supplier.
After 1973, when growth briefly stopped but Toyota kept demanding continuous price reductions based on continuous cost reductions, Toyota’s first-tier suppliers realized that they would need to reduce costs at their second-tier suppliers by teaching them the Toyota system. In this way, TPS trickled most of the way down the supply chain by the end of the 1970s.
Completing the Parallel Revolutions
As hard as it was to fully diffuse lean principles across Toyota’s physical production system, it proved even harder to complete the revolution in other aspects of the business. For example, Toyota Motor Sales gradually reduced its lead time to ten days for car orders from Toyota but it still retained a large bank of finished cars. It was not until Shotaro Kamiya finally retired from the chairmanship in 1981 (at age eighty-one) that Toyota could do the logical thing and merge TMS and TMC to form the Toyota Motor Corporation. After 1982, the inventory of finished vehicles in the Japanese domestic market withered to practically zero (before the collapse in demand after 1991 temporarily reversed the trend). 4 Most cars are now built and delivered within about a week of customer order. 5
Parts distribution long proved resistant to lean thinking, and Toyota did not apply lean techniques in its domestic service network (as described in Chapter 4 ) until the early 1980s. Up to that time, it was operating classic batch-and-queue warehouses even though the warehouses were supplied by the world’s leanest producers.
Finally, the initial shusa system Toyota put in place with the Crown in the early 1950s worked less and less well as the number of products began to proliferate. (Even as late as 1966, when the Corolla was launched, Toyota had only three automotive products, the Crown, the Corona, and its ill-starred “people’s car,” the Publica.) By 1991, Toyota was offering thirty-nine models of cars and trucks, based on nineteen separate “platforms” (auto-speak for the underlying body structure beneath the exterior sheet metal and interior trim).
The problem was that the initial, strong-willed shusa gave way to more bureaucratic personalities and Toyota’s functions became much deeper and stronger as the firm accumulated knowledge. The shusa, from their position deep inside the firm, had more and more trouble hearing the voice of the customer and often stumbled while pulling products through the development process. What was more, there was no adequate mechanism to inform shusa of each other’s work. As a result, many parts for new cars were being designed from scratch although almost identical components were either already available or were being developed simultaneously for other new vehicles. The results were excessive costs, a failure for more than a decade to reduce time-to-market (which became stuck at about forty-two months), and spectacular misreading of consumer desires as the Bubble Economy came to an end in 1991.
In 1992, therefore, Toyota reorganized its products into three platform groups (front-drive cars, rear-drive cars, and light trucks) overseen by truly heavyweight program managers with a much higher level of dedicated engineering resources. (The organization, in fact, now looks startlingly similar to Chrysler’s in North America, although Toyota would be reluctant to admit this.) The objective is to focus on product families which share components rather than on stand-alone products (each of which still has its own chief engineer), to dedicate engineering resources to the platform groups, and to streamline the flow of designs all the way into production so new vehicles can be carried from concept to launch in twenty-seven months. These are precisely the features of the product development systems we’ve seen repeatedly in our successful lean firms, except Toyota was late to adopt them.
Toyota Today
By the time we completed our previous book, The Machine That Changed the World, in 1990, Toyota had become the preeminent production organization in the world, and we believe this is still the case. Although the rules on data gathering for Machine prevented us at that time from identifying specific companies and facilities, Toyota finished first—and generally by a substantial margin even in comparison with other Japanese firms—on practically every benchmarking exercise we conducted—factory performance, product development time and effort (even before the 1992 reorganization), supply chain performance, and distribution. Surveys conducted since that time, as summarized in Table 10.1 , indicate that there has been considerable convergence in productivity and quality across the world but that Toyota and its parts group in Japan have retained their superiority.
Clearly, “the machine that changed the world” was Toyota’s interconnected ideas about product development, production, supply chain management, and customer relations systems. But pushing these concepts all the way through only one company, its suppliers, and its distributors, took thirty-five years. What’s more, even Toyota still occasionally wobbles in its course and the process of introducing lean ideas from one end to the other of Toyota’s value streams for its products is not complete even today.
T ABLE 10.1: R ELATIVE P ERFORMANCE IN A UTO A SSEMBLY AND P ARTS M ANUFACTURE , 1993–94
In the late 1980s, after Ohno and his generation left the company, Toyota began to consider the possibility that perhaps it should adopt more automation, indeed some of the aspects of high-tech mass production. The Tahara plant near Toyota City was the test case, where a much higher level of assembly automation was introduced with the launch of a new model in 1989. However, Toyota soon learned the same lesson as Roger Smith at General Motors: High-tech automation only works if the plant can run at 100 percent output and if the cost of indirect technical support and high-tech tools is less than the value of the direct labor saved. Tahara flunked both tests.
In its next plant, the Miyata facility on Kyushu, which opened in 1991, this lesson was taken to heart with a return to a much lower level
of automation in final assembly and a reorganization of the assembly line so that related activities—for example, the electrical system—are installed and then tested in one focused area. This gives the workforce immediate feedback on whether everything has been done correctly, a key factor in creating a psychological sense of “flow.”
Most recently, in the revamped Motomachi plant relaunched in 1994, Toyota has dealt with a key weakness of its system, the failure to evaluate the actual level of human effort involved in each production job and not just its feasibility within a given cycle time. By asking work teams to precisely determine the amount of fatigue and stress caused by each motion and then summarizing these for each job, Toyota for the first time can talk objectively about the level of effort required. This in turn permits the company to make jobs comparable (or to adjust the effort level for older workers or those with physical problems) and to answer critics who have frequently claimed that Toyota (and the Toyota Production System more generally) demand an impossible pace from workers. 6 If unacceptable levels of stress and fatigue are discovered, the work team then kaizen s the activities to redesign jobs and develop simple operator assist mechanisms.
Taking this step, involving a very considerable research effort, is a tacit acknowledgment by Toyota that for the indefinite future it will need about the same degree of direct human involvement in production tasks. The lights will still be on in the oft-predicted “lights out” factory well into the twenty-first century.
The new RAV4 vehicle for Motomachi also takes account of the fact that reducing the number of parts and simplifying their fabrication can be much more effective than either automation or a fast work pace in reducing product costs. For example, the body panels for the RAV4 take a maximum of three strikes to complete in the stamping shop, while panels in other Toyota models generally require five strikes. Going from five strikes to three automatically reduces tooling bills by 40 percent and increases the throughput of the stamping shop dramatically. Many other components in the RAV4 have been simplified as well. As a result, Toyota estimates that it has reduced the human effort needed to assemble a RAV4 by 20 percent, compared with the most comparable previous product, even while reducing the amount of assembly automation, the cost of production tools, and slightly reducing the work pace.
With regard to its total value streams, Toyota’s first- and second-tier suppliers all operate their production facilities in accord with the Toyota Production System, and have since the late 1970s. But the performance of the third-tier makers of small parts is still inconsistent. Some are good, some aren’t, and it remains to be seen if the latest yen shock is the crisis needed to push TPS all the way to headwaters of the parts fabrication value stream.
More striking, most raw materials suppliers (steel, aluminum, glass, and resins for molding plastics) are still stuck in the world of batch production. These firms, accounting for more than two-fifths of the total manufacturing cost of a vehicle, are outside the Toyota Group’s reach and most have been resistant to Toyota’s requests to streamline their thinking. For example, Japan has only three domestic glass producers, and until 1994 they were permitted by the Japanese government to operate a tight “capacity” cartel to control pricing and new entrants. Not surprisingly, one-month batches of pressed glass for automotive use have been the norm in the glass industry, and this seems to be typical for steel, aluminum, and plastic resins as well.
The magnitude of this problem for Toyota is shown by a simple calculation by Peter Hines of the Lean Enterprise Research Centre. 7 In the fall of 1994 he estimated manufacturing costs incurred along Toyota’s value streams as follows: Toyota itself, 22 percent; first-tier suppliers, 22 percent; second-tier suppliers, 10 percent; third- and fourth-tier suppliers, 3 percent; and raw materials suppliers (directly to Toyota and to each of these tiers combined), 43 percent. In the West, raw materials probably account for no more than 25 percent of manufacturing costs, but because Toyota has been so effective in cutting costs in its supply base through four tiers of suppliers while raw materials costs have not been managed in the same way, the real cost saving for Toyota today lies in changing the thinking and behavior of materials suppliers.
Finally, Toyota’s approach to aggressive selling was a great breakthrough in the 1950s but has hardly evolved since. The number of steps and the amount of effort involved in meeting customer needs through door-to-door selling creates a high-satisfaction/high-cost selling system when Toyota needs a high-satisfaction/low-cost system. Another leap will be needed (to be described in Chapter 13 ) if truly lean selling is to emerge at Toyota.
So even Toyota, the leanest organization in the world, has not yet succeeded in creating lean enterprises by removing all of the unnecessary time, effort, and error sources from raw materials to finished vehicle, order to delivery, and concept to launch, for each family of products. Part III of this book will propose ways to make this final leap .
The Spread of Lean Thinking Outside of Toyota 8
Because Toyota pioneered the full complement of lean techniques, it would seem that other Japanese firms should have been able to apply them much more quickly than Toyota. Yet, this was not the actual pattern. In the 1950s, Japanese electronics firms independently invented strong program management and a short product cycle, essential to their strategy of making a living by combining commodity electronic bits into clever packages and flooding markets with a variety of rapidly renewed products. However, only Mitsubishi, headquartered in nearby Kyoto and a member of the Chubu Industrial Engineering Association (of which Ohno was intermittently president), seems to have pursued Toyota’s experiments in production. 9
Other Japanese firms were making dramatic progress during this period, but along a complementary path and from a different starting point. They were steadily extending the original statistical quality control concepts introduced by the Americans immediately after the war 10 to involve the shop floor in Quality Circles using the seven quality tools and Deming’s Plan-Do-Check-Act problem-solving cycle. Soon they were experimenting with early forms of policy deployment and the management of quality improvement across each functional process. Within a few years, Total Quality Control (followed by Total Quality Management) was widely applied across the industrial landscape in Japan. 11
Stung by Nissan’s winning the Deming Prize in 1960, Toyota also began to adopt TQC in parallel with Ohno’s ideas and won the Deming Prize itself in 1965. By then, both quality and continuous flow were being managed as cross-functional activities reporting to the highest levels in Toyota. Toyota’s real advantage, as it turned out, was that it alone was able to combine TQC with TPS to stand out from others. 12
No one in Japan—even in the auto industry—seems to have paid much attention to Toyota’s unique approach until the first energy crisis in 1973. When most firms began losing money after years of steady growth, yet Toyota continued to earn healthy profits in a slumping market—by avoiding production of unwanted products and continuously pushing down costs—the virtues of Toyota’s lean system were suddenly apparent.
Mitsubishi Motors, which had already embraced many elements of the system, moved ahead rapidly on full implementation, and Mazda made TPS a central pillar of its comeback after 1974 (just in time for Ford to learn the system secondhand, beginning in 1979, when it took a 24 percent equity stake in Mazda). Nissan, Honda, and the other Japanese car companies began to do their homework as well, with mixed results. Nissan, to take the most striking example, found it very hard to give up on its own strategy of progressively automating activities to eliminate both human effort and the need for tight coordination by means of TPS. As a result, it fell steadily behind Toyota, after enjoying a comparable market share in the early 1960s.
Ohno realized that a major reason the Toyota system did not spread rapidly was that it required hands-on teaching. Yet no one with deep experience ever left Toyota except to go to a Toyota supplier. (Consultant Shigeo Shingo, who advised Toyota but who worked for many other firms as well,
was the one major exception.) Therefore, as Ohno contemplated retirement in 1978, he decided that a highly useful activity would be to take some of his most loyal and gifted disciples and form external propagation mechanisms.
The first of these, headed by his closest disciple, Kikuo Suzumura, was called New Production System, or NPS. 13 Ohno’s idea was to form a group of the chief executives from a range of Japanese firms outside the auto industry, including retailing. These were all firms selling directly to the public and none were competitors. They agreed to conduct hands-on improvement activities on the same model used by Toyota to spread TPS through its first-tier suppliers after 1969. Ohno was the “supreme adviser,” with Mr. Suzumura the day-to-day leader. As we have seen, Ohno also played a role in the formation of Shingijutsu as a more conventional consulting organization in the mid-1980s.
It seems fair to say that by the mid-1990s most major Japanese manufacturing firms and many of their first-tier suppliers were fully aware of lean concepts and most had at least some examples of implementation. However, we have been struck in our travels through Japan by the unevenness of implementation and the striking fact that many big companies placed their bets on the very different concept of high-tech mass production.
For example, we recently visited a large facility of a technically advanced firm where a race was being conducted between the rising yen and the elimination of expensive human effort. The process villages molding, cutting, and painting parts for the plant’s complex product were entirely automated, with robots neatly stacking the parts emerging from various fabrication steps on pallets to be taken by automated guided vehicles to an automated storage and retrieval center. From there, in-house parts and those received from vendors were taken automatically to a completely automated final assembly line which could instantly adjust its fixtures to hold any of the one hundred models of the basic product and assemble it solely through the efforts of pick-and-place robots. (The plant still employed 3,600, but none was involved in direct labor.) The facility exported 50 percent of its 7.5 million units of output and supplied one sixth of world demand for its product from one final assembly line in one room. For the future, this company is looking to China as a source of cheap subcomponents, currently supplied by local first-tier suppliers .
Lean Thinking Page 29