Lean Thinking

by Daniel T Jones

  When Showa’s original office and manufacturing complex in crowded Fukuoka City (on the northern tip of Kyushu, Japan’s southernmost island) was fully relocated in 1983 to new plants in suburban Umi and Koga, the management expected its fortunes to change. Instead, the decline continued. The production system in the new plants was in fact the same as that in the old. Process villages for casting, cleaning, stamping, welding, painting, and assembly were run in batch mode with long intervals between tool changes. This practice created mountains of parts which were then taken to central stores before reshipment to the next processing step. Orders took months to work their way through the system, as chased by expediters with hot lists. (It was the familiar world of every firm we’ve looked at before the advent of lean thinking.) In addition, the cost of starting exports was high and the diversification into ornamental castings pitted Showa against larger firms with established reputations in the building trades.

  It was at this point that Tetsuo Yamamato decided he needed to take dramatic action. He would contact Taiichi Ohno and ask for help.

  This was not a trivial decision, because Ohno’s reputation was one of unrelieved ferocity. He was barely able to suffer geniuses, and the fools he seemed to find all around him could expect routine tongue lashings for failures they scarcely understood. (Chihiro Nakao, one of Ohno’s favorite pupils, worked with the master for more than twenty years but cannot remember ever receiving a compliment of any sort from Ohno for his efforts. He can, however, remember receiving tongue lashings, almost by the day.) What was more, Ohno might not even be available. To date, he had not formally agreed to help any firm outside the Toyota group.

  On the other hand, Ohno was clearly a genius—one of the preeminent industrial thinkers of the twentieth century—who had transformed the Toyota group into the most competent manufacturing organization in the world. If it was only a matter of putting up with insults, Yamamoto felt the price would be worth the reward. In addition, as a man of Ohno’s generation who was president of a golf club in the Fukuoka area and a mah-jongg master, Yamamoto thought he could entice Ohno beyond the Toyota group with an ample supply of his two favorite entertainments. Perhaps in the process he could divert Ohno’s scorn from Showa employees.

  When Ohno accepted an offer to address the Fukuoka Chamber of Commerce late in 1983, Yamamoto acted as his host and seized the opportunity to invite Ohno to return early in the new year for a round of golf and a quick look at his foundry. Fortuitously, Ohno at just this time was pondering what to do about some of his lieutenants, including Yoshiki Iwata at Toyoda Gosei and Chihiro Nakao at Taiho Kogyo. He was getting old and they feared that once he was gone they would be penalized for his famous clashes with his peers at Toyota.

  These had occurred repeatedly during Ohno’s take-no-prisoners campaigns to push the Toyota Production System through Toyota itself in the 1950s and 1960s and then, after 1965, through the supply base. After the conversion of the first- and second-tier suppliers was largely completed in 1978, Ohno was no longer so critical to Toyota, and he had been eased out as executive vice president. His new jobs as chairman of Spinning and Weaving and of Toyoda Gosei sounded impressive but were, in fact, largely ceremonial, designed to give him recognition for his past achievements but to keep him at a comfortable distance from Toyota Motor Corporation at the heart of the Toyota group.

  The Showa invitation suddenly raised the possibility of solving several problems. It would give Ohno a continuing testing ground for his ideas in a firm completely outside the Toyota orbit, one mired in the classic world of mass production, and it would provide a chance for some of his loyal deputies to leave the Toyota group to form a consulting firm, to be called Shingijutsu, for “new technology.” (As we will see in a moment, he had already had a parallel idea for another organization, called NPS, or New Production System, which he started a few years earlier with other loyal disciples.) So, Ohno took a look at the Koga foundry, roared his famous roar, and then quietly said “yes,” he and his associates would take on the leaning of Showa Manufacturing.

  The Initial Struggle

  We have encountered many Americans and Europeans who seem convinced that lean thinking somehow comes naturally to the Japanese. (These same individuals routinely assume that all Japanese firms are lean and have been for decades, another notion which is completely wrong, as we’ll show in a moment.) The reality is better represented by the initial reaction of the workforce when Ohno and his colleagues started their first improvement activities at Showa’s foundry.

  Ohno immediately asserted that by moving to small-lot production and producing only what was requested by the next production step, it would be possible to reduce the three months of inventory of the typical part to a few days. Time-to-market could therefore be reduced to a fraction of current levels. He stated that it would also be possible to double labor productivity, halve the amount of plant space needed for current output, and to do this very quickly with practically zero capital investment. (These are numbers the reader will no doubt recognize as “normal” in a lean transformation.)

  The Showa workforce, however, was completely skeptical and resistant. They were mostly longtime foundry workers, and they simply “knew” that none of these objectives was achievable except, possibly, by means of their working much harder. And the line management felt little different. Plant manager Kawabe, for example, was still smarting from his initial encounter with Ohno and believed that techniques right for the high-volume auto industry were out of place in the low-volume casting and boiler-building business.

  Nevertheless, because Ohno and his disciples had President Yamamoto’s full backing, it was at least necessary to go through the motions. The first project, as shown in Figures 10.1 and 10.2 , was to convert coil making and assembly from a batch process to single-piece flow by creating a cell for the pipe-cutting, fin-press, expansion, cleaning, brazing, leak testing, and final assembly steps. High-speed machines that were hard to change over were replaced by designs created in Showa’s tool shop (eventually totaling three hundred throughout the company), so that the cell could convert from one coil design to another in only a few minutes before resuming operations. The output of the cell was then fed directly into a simplified and shortened final assembly track.

  Despite the skepticism of the workforce, and pointed disagreements with them at almost every step, in less than a week it was possible to eliminate half the plant space, 95 percent of the in-process inventory, half of the human effort, and 95 percent of the throughput time needed to make a coil. (In addition, quality improved dramatically.) The capital investment and time needed for the transition were trivial in comparison with the benefits.

  These were electrifying numbers in an old-line organization like Showa, which had had hardly any productivity growth in decades. Yet they were exactly what Ohno had promised. As the kaikaku campaign progressed from activity to activity, substituting single-piece flow for batch-and-queue, the results could not fail to gain the attention of even the most negative among the Showa workforce. As attitudes began to shift, Takeshi Kawabe—the most skeptical of the original management—was even willing to take on a new job as head of the newly created Production Research Department. (Elsewhere, we’ve seen this new function called the Process Improvement Department [Lantech], the JIT Promotion Office [Wiremold], the Continuous Improvement Office [Pratt & Whitney], and the GROWTH Division [Freudenberg-NOK].) He took charge of improving every activity across the firm and gradually became the in-house Ohno.

  F IGURE 10.1: C OIL M AKING AT S HOWA , S PRING 1984

  F IGURE 10.2: C OIL M AKING AT S HOWA , S UMMER 1984

  Over the next three years, as Kawabe 1 exhibited the enthusiasm of the new convert, every activity was rethought and improved at least once. And eventually, in pursuit of perfection, every activity was kaizen ed at least ten times. Productivity soared, inventories were slashed to a quarter of their former value, and the amount of space needed to make a given amount of output was cut b
y 75 percent, as shown in Figure 10.3 .

  F IGURE 10.3: S ALES , P RODUCTIVITY , S PACE U SE, AND I NVENTORIES AT S HOWA , 1984–92

  Source: Showa Manufacturing Co., “Background of Implementation of the Showa Production System,” 1993, p. 5.

  As a result, Showa clawed its way up from deep losses to modest profits. Still, selling prices for Showa’s products continued to decline in a stagnant market. Showa had bought time to think, but it was clear that cost cutting alone would not be sufficient to generate adequate profits.

  A Contradiction in Thinking

  A key problem—one faced by many Japanese firms today—was that Showa’s market strategy was at odds with its new production methods. Showa had discovered how to build a complete boiler in four days (compared with sixteen to twenty weeks) and how to build all boilers to special order without paying a significant production cost premium, yet the firm was looking to overcome weaknesses in its Japanese markets by selling standardized products in the American market at the end of a three-month distribution line. At such great time and distance, no customization or rapid market response was possible. What was more, the export drive had hardly gotten into full swing before the yen began to strengthen steadily, soon doubling in value from 260 yen to the dollar in February 1985 to 129 in February 1988.

  Clearly something was wrong when a highly flexible firm was looking desperately on the other side of the world for standardized business, so President Yamamoto launched a rethink of Showa’s whole strategy and product line. He concluded that there was simply no future beyond replacement demand in Showa’s traditional product line of cast-iron boilers, even if some competitors could be forced out of the business. (Remember that to keep his core workforce busy and reap the full financial benefits of the lean transition, he needed to double sales very quickly at constant prices.) He also concluded that profitable exports through long supply lines were a mirage.

  Yamamoto therefore decided that Showa should reverse course and work backwards, asking what its key technologies and capabilities really were and how these could be matched up with new needs among domestic consumers. As he looked at the booming Japanese economy it seemed obvious that the Japanese were underspending on themselves—both in their public goods and in their private lives. Therefore, the most promising growth opportunity would be to build lower-volume, customized goods supporting a new and higher-quality lifestyle for domestic customers. However, Showa’s functional organization was ill-suited for this new task.

  A New Organization to Support Leanness

  In 1987, Yamamoto broke up 104 years of centralized corporate structure by creating new, horizontal product teams, one each for a range of new product lines. These product families eventually ranged from custom-designed and strikingly original cast parts for “showcase” bridges (for example, in public parks) to low-volume air-conditioning units for specialized applications. Other business units were created for custom-designed truck bodies for the construction industry, special aluminum castings—practically sculptures—for public buildings, and custom castings in exotic alloys for the aircraft engine and nuclear power industries. A particularly important initiative was an “environmental products” unit making home air filtration systems and home bath heating and filtration systems to keep the water in the tub hot and clean twenty-four hours a day. (One business unit established to manufacture the automated parking carousels tucked away in the back of most Japanese apartment buildings failed and was eliminated.)

  Each product team had its own marketing, product design/engineering, and production system, renting space in Showa offices and plants as appropriate. Within a brief period the centralized, “batch” operations of the old Showa—marketing, design, and production—were eliminated and replaced with dedicated, continuous-flow teams for each product family. These employed a very high fraction of Showa’s total headcount. Only a few workers were left in the tiny, centralized functions consisting of production scheduling, finance, supplier development and logistics, human resources, quality assurance (to deal with complaints from customers), and, of course, “production research” to continually improve every activity.

  In the new system, a high fraction of costs were directly assigned to individual products and only a small fraction were allocated from general overhead, so it was possible to know whether product families were producing an adequate profit. As a result, the leaders of each product team could easily be judged on their bottom-line success. The team leaders were told to continuously renew their product ranges and to be prepared to exit product lines where they could not make money.

  Between 1984 and 1995, Showa replaced 100 percent of its product range. In the process, it eliminated two thirds of the products and production tasks which had been so carefully and repeatedly kaizen ed. Showa’s current president, Keiji Mizuguchi, notes that rapid entry and exit from product lines is “normal” in a world of custom products but would never have been possible in the centralized organization of the pre-1987 Showa. Nor would anyone have known which products were making money and which were dragging down the firm.

  From Hard to Soft Kaizen

  The objective of each product team was to introduce single-piece flow in product design, order-taking, and production—the same approach taken by Freudenberg-NOK, Lantech, Wiremold, Pratt & Whitney, and Porsche. Because the production steps were soon all kaikaku ed (then kaizen ed and re-kaizen ed), it was gradually both possible and appropriate for the Production Research Department to move beyond the plant to help rethink product development and order-taking.

  The first step, beginning in 1991, was to rethink the already streamlined design process to take full advantage of Showa’s commitment to customization. Clearly, if boilers, bridge railings, and ceilings for shopping malls were going to be customized, the customer needed to be directly involved in the design from the outset, but Showa, in far-off Fukuoka, had no easy technical means for doing this. Therefore, Takeshi Kawabe (who only seven years earlier had been the manager of a classic batch-and-queue foundry) undertook a three-year project to develop an interactive design software which the customer and Showa designers could look at together in real time to make decisions about product specifications and the state of orders. This was introduced in 1994.

  At the same time, Showa rethought its boiler technology and materials, switching to stainless steel and to new production tools designed in-house which eliminate the need for workers to weld inside the boiler vessel. Using the new design method and the new production system, boiler costs were reduced by an additional 30 percent in Showa’s most mature and problematic product family.

  The Final Element: Rethinking Order-Taking and Scheduling

  By the time Tetsuo Yamamoto retired as president to become chairman in 1993, Showa had nearly completed its transformation from mass to lean producer. The major organizational step left for new president Keiji Mizuguchi (who came from the giant Sumitomo Trading Company, which handles the distribution of many of Showa’s products) was to rethink order-taking and scheduling. In doing this he was inspired by the American reengineering movement, but in the end he went further.

  As Mizuguchi looked at the situation in 1993, Showa was able to physically build almost all of its products in less than a week. Yet it was accepting orders months in advance, particularly in the building industry, where many of the items needed to complete a project really did require months to fabricate in other, mass-production firms. Part of the problem was that customers were constantly changing their orders right up to the last minute. What was more, Showa was running all of its orders through a centralized Production Scheduling Department which processed orders (and changes) in batches before sending them along to the design and production groups in each business unit. Due to time pressures (because orders required several weeks to process) and the many hand-offs from one department to the next, orders were sometimes started into production that were clearly nonsense—impossible specifications, for example—creating the need for expensive
rework.

  A simple approach would have been to create a streamlined order scheduling department with multiskilled workers to handle orders one at a time and see them through the system. However, this method retained the centralized Scheduling Department and Mizuguchi concluded that this was not lean enough. Instead, the reengineering team eliminated the Scheduling Department and gave the task of scheduling orders to the marketing group in each product team.

  The product teams were told to schedule backwards (working to takt time) to precisely synchronize orders with available production slots at a point exactly four days before shipment when the firm order needed to be inserted in the production schedule. This is exactly the system used at Lantech, as described in Chapter 6 .

  In this new system, orders with incorrect information must never be passed forward by the designers and engineers. (Scheduling equivalents of poka-yoke devices have been developed to make sure all mistakes are caught.) Meanwhile, the customer must be educated to understand that Showa needs only four days of lead time before the product is ready to ship so that there is little point in specifying exactly what is wanted (and then changing the order repeatedly) until it is time to build it. The customer must also be educated, as at Lantech, about the curious fact that Showa now ships exactly on schedule.

  The final element of the Showa ordering and scheduling system is that it is completely open for everyone along the value stream to see—the customer, the distributor, the Showa product team, and the component and materials suppliers. Only the product team can change the information on the electronic schedule board, but everyone with an interest in the outcome can electronically check on the status of orders at any time. Another example of the power of visual control.