At the end of the week, the initial rundown in inventory was complete (there was no longer anyplace to store twenty-eight days’ worth of parts) and the effects were both dramatic and completely visible. The Porsche internal teams, meanwhile, had made hardly any progress on their parallel tasks and concluded that they should simply join the next consultant-led kaizen.
Many improvement activities lay ahead in engine assembly, as shown in Figures 9.3 , 9.4 , and 9.5 , tracing the transformation of engine assembly between the fall of 1992 before the transformation started to the end of 1993 when a fully lean system was in place. Over this period, the amount of space for inventories was reduced from 40 percent of the assembly area to zero, the amount of parts on hand was reduced from twenty-eight days to essentially zero, and parts were in the assembly area for only about twenty minutes before the completed engine was sent to the final assembly area.
F IGURE 9.3: P ORSCHE E NGINE A SSEMBLY , O CTOBER 1992
Instead, parts kits for each engine were built up in a kitting area on the floor below and sent up to the assembly floor in little carts at exactly the rate engines were being assembled. (The kits were themselves a poka-yoke device because the parts were placed on the cart in their exact assembly sequence. Any part skipped over would be spotted immediately.) 11 Meanwhile, a kanban system was being installed with major suppliers so that the needed parts were delivered directly to the kitting area at frequent intervals. The massive automated central warehouse Porsche previously used for received parts was partly emptied and the space made available to the service parts organization.
F IGURE 9.4: P ORSCHE E NGINE A SSEMBLY , D ECEMBER 1992
F IGURE 9.5: P ORSCHE E NGINE A SSEMBLY , D ECEMBER 1993
At the same time, improvement activities were started in the paint booth, the body welding shop, the engine machining shop, chassis assembly, and final assembly. On their monthly one-week visits, the Japanese consultants would oversee the efforts of all six improvement teams beginning with an analysis session on Monday morning and a report to all six teams in the afternoon on the proposed plan of attack.
Because they had invariably seen the same situation before—remember that they and other Japanese sensei have been conducting similar exercises every week for nearly thirty years—they could instantly point out opportunities for additional improvements going beyond what the team had initially proposed. As Wiedeking commented, “You have to actually apply lean thinking in real situations to learn to see. Nakao and our other advisers have developed 20/10 eyesight, so we could all learn at a multiple of our normal speed. It was astonishing.”
With the six plans agreed upon, the teams went to work—senior managers, production workers, support staff—to build any necessary equipment, move machines, run the new layout, standardize the work, and stabilize the whole activity. It was generally possible to continue production while conducting the improvements because machines could be moved in the evening or over the lunch hour. By Friday, it was time to summarize the improvements, hear the reports of all six teams, make a list of follow-up activities required to sustain the improvements (often very long), and celebrate.
Gradually, over a two-year period, the PVP teams that planned and followed up on the consultant-led workshops gained the experience to join with work teams and conduct activities without outside help. A policy was then adopted that every work team would conduct a major weeklong improvement project on its activities every three months, in addition to taking immediate action on improvement suggestions from work team members at any time. These activities in turn became the key to meeting the measurable improvement targets for every work team set as part of the Porsche Improvement Process.
Dealing with the Jobs Problem
Wiedeking would not have gotten very far if he had not faced up to the jobs problem. Part of the problem was addressed by the prior decision to bring assembly work back in from Audi for the 968 model. Another part could be handled by contracting with Audi and Mercedes to assemble a few of their ultra-low-volume models. And part of it could be handled by assigning excess workers with special skills to kaizen activities for extended periods. For example, in the paint booth, some of the skilled paint finishers were assigned to improvement teams trying to eliminate contamination from the system by finding the root causes so less end-of-the-line touch-up would be needed. When volume picked up again (as it had to if Porsche was going to survive at all) these workers would be needed once more for painting.
However, production at Zuffenhausen was falling from 26,000 in 1991 to 14,000 in 1993 and it seemed unlikely that it would return to the 1980s levels for years, until the introduction of new models. In addition, it was apparent that Porsche was engineering and making a wide range of parts and components in-house at absurdly low volumes and high costs. These needed to be bought instead from the firms supplying similar parts to the big car companies. Therefore, it was apparent that Porsche simply had too many people to survive.
A onetime adjustment in the workforce of twenty-five hundred employees was carried out over a three-year period beginning in mid-1992 to bring the headcount to a level consistent with long-term needs. Some workers took a special retirement offer and others were given a large severance. Because natural attrition is about 3 percent a year, given the age distribution of Porsche’s workforce, an additional 30 percent reduction in the workforce can be achieved in the next decade without resort to layoffs if no additional sources of production volume can be found.
While this reduction in headcount was taking place, the management offered the standard guarantee we’ve seen in all of the examples cited in this book. It made a commitment to the works council that no one would ever lose their job due to the introduction of lean thinking by means of periodic PVP activities, although the nature of everyone’s job would constantly change and a collapse in sales might necessitate another round of departures to save the company. This guarantee was originally given for the three-year period 1991–1993, and later extended for another three years through 1996.
The Reaction of the Workforce and the Union
Both the workforce and the union were initially quite upset at the affront to, respectively, their competence and their role. The lean message was that the traditional craftsmanship was mostly muda: correction of mistakes which should never have been made, movement to find parts and tools which should be immediately at hand, wasteful motions through a lack of careful analysis of how to do the job, wasted time while watching machines which could be taught to monitor themselves, waiting for missing parts, and inventories everywhere due to batch-and-queue methods.
Another aspect of the message was that the Works Council should participate directly with the management in problem solving by participating in improvement activities. A hands-off, reactive attitude which implicitly assumed that jobs and living standards could be preserved simply by bargaining to extract them from management was simply irrelevant to the new situation of the German economy.
Fortunately, lean thinking carries a positive message which can redefine craft for a postcraft age. As Porsche employees participated in one improvement activity after another, many began to see that there is a higher form of craft, which is to proactively anticipate problems in a team context and to prevent them while constantly rethinking the organization of work and flow of value to remove muda. (Another way to think of it is that Chihiro Nakao is the ideal-type craftsman for the twenty-first century.) Thus the direct worker and the work team subsume many of the traditional activities of “management” while improving activities at a far more rapid rate than management alone ever could.
The special strength of a firm like Porsche in this respect is that the workforce is highly skilled in the fundamental disciplines of manufacturing operations. Multiskilling, job rotation, analysis of root causes, preventive maintenance, and kaizen are all more productive activities for a workforce with these skill sets, and Nakao was soon complimenting the improvement teams on coming up with ingenious stratagems which even he had no
t thought of. (We’ll have more to say about this in a moment.) In short, Porsche was and is still a craft company, but the craft is becoming the new lean craft of rapid and radical continuous improvements.
Fixing the Supply Base
Because Porsche buys nearly 80 percent of its manufacturing value from suppliers and is increasing this fraction, it was immediately apparent that teaching the suppliers to see was as critical as teaching Porsche employees. A number of suppliers had recently agreed to just-in-time deliveries, but when Porsche personnel investigated, the suppliers were invariably supplying just-in-time from massive warehouses. The demand for frequent deliveries in small batches had had no effect on production methods for the simple reason that most suppliers had no idea how to perform small-lot production.
Teaching 950 suppliers to see was clearly hopeless given Porsche’s resources and the small fraction of most suppliers’ output bought by Porsche. So the first step was to start reducing the supply base to 300 firms, partly by standardizing many parts and dropping low-volume options. Within this group of 300, about 60 suppliers were designated as critical systems suppliers and it was often possible for the former direct suppliers to become second tiers to these firms.
Porsche then formed a supplier improvement team, called the POLE team (from the racing term for the lead position at the start of a race), with the objective of obtaining the “pole position” in the race for survival. The team proceeded to conduct the exact same improvement exercises at the most important suppliers as were occurring inside Porsche. They began with those suppliers most receptive to lean thinking, like seat-maker Keiper Recaro, and used the initial successes with these firms to encourage the more reluctant suppliers to join in. The objective was to get material to flow continuously through the suppliers as Porsche pulled it while at the same time dramatically reducing both the number of defective parts (running at 10,000 per million in 1991) and the need for Porsche to assign a hundred workers to incoming inspection.
The experience was always the same. As Manfred Kessler, the director of the POLE team, remembers: “When we arrived at supplier plants, the management would always insist that there was nothing to improve. They’d wearily say, ‘We’ve rationalized everything already in response to visits from the supplier development teams from other manufacturers. There was really no need for you to come.’” The POLE team would then ask the senior management to play the Porsche version of the JIT game, 12 a simple exercise in which five senior managers take roles in a four-stage production process folding and packing three colors of paper boxes. (The game is described in Figure 9.6 .)
The first person is asked to bundle up and deliver quantities of unfolded boxes in three colors to the two pre-assembly stations. The quantities are in response to a customer order. One pre-assembly station folds the large boxes while the other pre-assembly station folds the small boxes and both stations secure their boxes with a rubber band. The boxes are then passed ahead to the assembly station where the fourth player opens the large box and places the small box inside. The player writes out a ticket, folds it, places it on top of the small box, and then closes the large box and secures it with a rubber band. The box is then passed to quality control/dispatch where the fifth player opens the large box and checks to see that the ticket is present and properly written. This player signs and stamps the ticket before placing it back on top of the small box. The large box is then closed, secured with a rubber band, and delivered to the customer.
The players are told to work at their own pace to produce the three colors of box in response to the customer order. Soon every player is trying furiously to complete his tasks, first for one color of box, then for the next. However, a huge mountain of boxes quickly builds up in front of the fourth player, who has a bigger job than the others. In addition, the customer announces that he wants to change his order, to receive first whichever color of box the team has left till last. This quickly produces even more of a pile-up as the wrong color boxes are pushed to the side so the right color can get through .
F IGURE 9.6: T HE P ORSCHE JIT G AME
The team of five is then asked what’s wrong and what could be done about it. The answer is always the same: “The fourth player is the bottleneck so we need to add another worker to the assembly step and build a storage area between steps two and three.”
The POLE team then suggests that instead the five players should try a pull system by making only five boxes at a time and only when asked (pulled) by the next player downstream. To the players’ amazement, the whole activity proceeds smoothly, with only a tiny buildup of boxes between steps two and three. They then play two more rounds, reducing the lot size to three and then to one, eventually achieving perfectly smooth flow and no buildup of boxes at all.
Next, the POLE team says the customer is going to vary his order at random between the three colors of boxes and asks what will happen. The supplier executives recognize this situation as the key headache in their lives and predict chaos. But, of course, with no boxes piled up in inventory, it’s a simple matter to switch from one color to the next.
As the supplier’s managers are scratching their heads, the POLE team moves from games to reality by suggesting that the exact same techniques should be introduced in the activities required to make Porsche parts. “Why don’t we take a set of activities for one part and try it today?” The POLE team would then stay for a week or two to remove all the waste the team could find and to standardize the process and develop follow-up steps so the new level of performance could be sustained. The understanding with the supplier’s management from the outset was that the cost savings would be precisely calculated and divided three ways, one third to the supplier, one third to Porsche, and one third as a pass-through to the Porsche customer.
On some of the toughest cases Nakao was brought along for shock value, but in general the Porsche team was able to do the work itself, and invariably with the same results: a halving of effort, a 90 percent reduction in throughput time from raw material to finished part, the complete elimination of in-process inventories, and a dramatic improvement in quality. At the end of two weeks of full-time work by the six-member team, when the full effect had been demonstrated—generally to widespread astonishment—the POLE team specified that any Porsche systems supplier must develop its own POLE team and go through the muda elimination exercise for every part supplied to Porsche. Then, of course, it ought to go to work on its own suppliers.
After two years of full-time effort at the end of 1995, Porsche had conducted muda elimination exercises of several weeks’ duration at the plants of thirty of its sixty largest suppliers and had worked with a few second-tier suppliers as well. Because of inquiries from many companies not supplying Porsche but who had heard about this activity, Porsche has now started Porsche Consulting, an external consulting practice similar in concept to that started by Freudenberg-NOK in North America. Thus, Porsche is not only a world-class product-technology consultant but hopes to become a world-class consultant on lean thinking.
Fixing the Overall Management
As the results of the conversion to lean thinking began to appear in physical production, something began to happen which we’ve seen in other companies (for example, Lantech and Pratt & Whitney). Power began to shift from the product engineers who had dominated the firm for its entire history to the operations managers. The supervisory board suddenly noted that what was dramatically improving was physical production, an activity not previously thought to be central to the firm’s success. The results were especially spectacular in terms of inventory reductions, which were freeing up the cash Porsche badly needed to fund its new product program.
The supervisory board therefore took a step which previously would have been unthinkable at Porsche. Wendelin Wiedeking, the operations director, was promoted to chairman of the management board and exhorted to apply the same medicine to the entire company .
On taking office in August 1992, Wiedeking shifted all of the senior
managers to new positions and persuaded many to retire. He was certain that they were anchor-draggers whose long experience and intense loyalty to Porsche as it had been would prevent them from ever embracing a new way of thinking.
Fixing the Product Plan
The most critical first step was to straighten out the model strategy. The plan had been to withdraw from the mid-price range and to field only ultra-high-performance cars, larger and fancier than the 928 model, on the presumption that Porsche could not offer a mid-priced driver’s car with reasonable performance and make a profit. However, it now appeared that costs could be dramatically slashed and that the segment retreat strategy being pursued would pit Porsche against BMW, Mercedes, Audi, and eventually the Japanese.
Wiedeking decided that Porsche must concentrate entirely on the niche it had itself created (“Let’s make originals, not copies!” he pronounced) and produce two new two-seat sports cars for this niche at different price and performance levels but sharing about 40 percent of their parts, including the engine block, to make the plan feasible. These are the mid-priced Boxster, introduced in the autumn of 1996 to replace the 968, and an upscale successor to the 911, introduced in 1997.
Because this is clearly a niche with limited volume, the second part of the product strategy was to take on the task of developing and building low-volume coupes, cabrios, and even luxury vans for the big German car companies. (Several projects are under discussion.)
It’s impossible to know whether this bet-the-company decision is the right strategy, although this will become clear just as this book appears. It is a very clear strategy which vaporized the paralyzing confusion in the company about what a Porsche “is.”
Fixing the Product Development System
The success of the new strategy is unavoidably a roll of the dice beyond Wiedeking’s control. What Wiedeking could control was the method of developing the new cars so they really are superlative additions to the long Porsche tradition of slightly tame drivers’ cars, but created in the least time with the lowest feasible engineering, tooling, and production costs .
Lean Thinking Page 25