IMPROVING THE SOUL OF AN OLD MACHINE
Across industrial America, reengineering teams are trying to bring factories up to world-class standards. James B. Treece, who covers manufacturing for BUSINESS WEEK, joined one effort at an Indiana plant. Here's his report:
It's hot, dirty, and hectic inside the Freudenberg-NOK auto-parts factory in Ligonier, Ind. On the engine-mount line, workers lean over searing molds, loading and unloading raw materials and finished parts. Laboring three shifts a day, they produce 123,000 parts a month for Ford Motor Co.'s Aerostar vans and Ranger pickup trucks. Like many U.S. factories, however, this 8-year-old plant has plenty of room for improvement. On this late summer day, our 12-member team has been given ambitious goals: boost capacity by 20%, raise productivity 15%, and reduce work-in-process by 50%.
Obviously, we can't produce plantwide gains that dramatic in the three days we have to come up with suggestions. Instead, our group of plant workers, managers, and outsiders--including one reporter--is one of a series of quick-hit teams trying for improvements in one section of the plant. Between Monday and Thursday morning, we'll scour the engine-mount line, proposing and testing changes to reach our goals. If our ideas work, they'll be spread to the rest of the factory. Then, other teams will move in, and the process will begin again. In the next year alone, 40 such teams will move through this sprawling factory on the edge of a cornfield in the Amish countryside of north central Indiana.
SWARMS. These are the front lines in the battle for supremacy in world manufacturing. Particularly in the auto and electronics industries, lean-manufacturing teams are swarming over America's factories. Their aim is to boost productivity by bringing what the Japanese call kaizen--continuous improvement--to the shop floor. To do that, the teams are introducing a blizzard of small, inexpensive changes that should make U.S. plants more competitive. "What we are doing today is what any company will have to do to survive a decade from now," says Joseph C. Day, chief executive of Freudenberg-NOK, a $420 million German-Japanese joint venture with 14 U.S. plants that make rubber seals, vibration dampers, and engine mounts for securing the motor to a car's frame.
Although his industry isn't known for manufacturing innovations, Day, 48, has embraced kaizen. In 1992, he began dispatching teams into his plants to bring them up to speed after seeing how well lean production worked in the company's new factory in Cleveland, Ga. He believes the groups--dubbed GROWTTH teams, for Get Rid Of Waste Through Team Harmony--will help the company more than double sales by 2000, to $1 billion--without adding people or factory space. They also provide a way to respond to demands from customers, such as Ford and Chrysler for price cuts of 6% or more a year. The teams already have been through several operations in our plant. Day has smoothed the way by promising that no one will be laid off as a result of their suggestions. Otherwise, he figures, good ideas will be hard to find.
It has been less than three months, in fact, since our engine-mount line last went through the kaizen process. Back then, a GROWTTH team combined what had been a two-person press-and-paint operation into cells of a single worker each. Workers were promised that a follow-up group would try to ease their heavier workload. That's us, says team leader Jerry Decker, a plant quality manager. We're also looking for ways to meet today's orders using five presses instead of the current six. That will free one for a new engine-mount order the plant expects to win next spring.
After a morning training session, we grab our stopwatches and white lab coats emblazoned with the GROWTTH logo and head onto the floor. Towering above us are five three-story-tall presses, which melt 100-foot-long strips of natural rubber, inject part of the 235F molten mass into a series of 4-inch-by-4-inch molds, and quickly cure it. At each press, a worker buffs excess rubber off the latest batch of steaming engine mounts with a spinning wheel of metal bristles that resembles a power sander, then does two other operations. As we watch, the long strip of rubber being fed into the press breaks. As we time every step, the machine operator sprints up a short stepladder, reloads the rubber, hustles through several other tasks, then waits for the press to open again.
DIRTY WORK. When we reconvene in our conference room, we agree that the workers, who finish their tasks in about 13 minutes, are not the problem. The bottleneck is the 18-minute machine cycle. That can be improved without overtaxing the workers, Decker believes. So he plans an experiment. Can we shorten the cure time from 14 minutes to 12 minutes without affecting quality? By the next day he has his answer: Yes.
At 5 a.m. on Tuesday, headquarters-based salesman Rich Hamed and I arrive at the plant to ask the third-shift crew about their jobs. Byron Burns gives us an earful. He has been up and down the short ladder 10 times since 1 a.m. to refeed broken rubber strips. He shows us his left arm, blackened from overspray at the paint booth, then pulls up his blue work shirt to reveal tiny black dots covering his belly. Rubber ripped off by the buffer flies through his shirt and into his pores. "I stand in the shower," he says, "and pop these out." Workers long ago removed a safety shield, after it got so blackened they couldn't see through it.
At breakfast, Decker parcels out eight key issues to subteams. One will focus on loading and unloading the press. I volunteer for a team that will ease the workers' duties and deal with Burns's woes. My group includes line worker Shirley Voirol, Rick Goslee from maintenance, and Forrest Knox, an engine-mount buyer from Chrysler. We head back into the plant to confront the buffer. The spinning metal brushes are a hazard, and workers risk carpal-tunnel syndrome from forcing the mounts against them with their wrists at an angle. The previous kaizen team had a prototype of a semi-automated buffer built but never got it to work right. We start to tinker with it, a task that will consume the next two days, and compile a list of nine suggestions--such as one for eliminating the workers' long stretch down into four-foot-tall parts bins.
As the teams go over their lists, Decker tallies his computations. He figures the plant already could meet demand with just five presses, even without working Saturdays. But that assumes no unscheduled downtime. No presses idled, as one is today, waiting for replacement parts. No 45-second holdups when the rubber breaks. No loss of a minute or two when a worker, busy elsewhere, doesn't realize the press is ready to be unloaded. These small problems add up. Says Decker: "We're being nickeled and dimed to death."
The next day and a half is chaotic. Subteams test and retest ideas. The problem of breaking rubber strips is solved by feeding them off a large cable spool, instead of yanking them from boxes. Warning lights are added so workers facing away from the presses will know when the machine cycle is finished. Other problems are more intransigent. A group tries to reduce excess spray in the paint booth, but the paint gun keeps clogging. Now and then, Knox or I suggest automation as a solution to a problem, but we're ignored. The mandate is for quick, cheap solutions, not major capital spending.
As the changes accumulate, we bring out a cutting torch and alter the work- cell layout. We move parts bins, install a paint shield, and yank out old, unneeded equipment. By adding an air jet to cool the mounts' surface so workers don't have to wait as long to paint them, we halve in-process inventory. In all, we cut the distance each worker must walk per cycle from 32 feet to 17 feet.
By late Wednesday afternoon, we recognize that some projects will have to be left for later. "We identified a lot to do," says Decker. "The real kaizen is the next 30 days." To make sure unfinished business gets done, a former plant manager, Bren Harris, will oversee follow-up on our six-point "to do" list.
PRIME CULPRIT. Thursday morning, we present our results to Michael J. Crosby, senior vice-president and general manager of the Rubber Products Div. We have met two of our three goals: We cut work-in-process inventory by the targeted 50% and improved productivity 23% on the Ranger line and 17% on the Aerostar line--reflecting the different cure times of different-sized parts--against a goal of 15%. But we upped capacity only by 7.5% on the Ranger line and 11.6% on the Aerostar line, short of our 20% goal. Still, we identified the prime culprit that keeps the plant from operating at its peak: unplanned downtime, which was running significantly higher than plant officials realized.
There's no feeling of conquering-hero camaraderie. Perhaps we're aware of how much remains to be done. Nonetheless, a modest investment of time and money has paid off dramatically. A month after I leave, all but one of the changes we recommended have been put in place. With nearly a dozen other plants going through a similar process, Day estimates that by the end of 1995 he will have invested $2 million on the teams and their suggestions companywide--for a direct payback of $12 million to $20 million. With more American manufacturers adopting kaizen techniques and realizing similar returns, a little investment should go a long way toward reinvigorating the U.S. industrial base.MAKING KAIZEN PAY OFF
Kaizen--a Japanese term for "continuous improvement"--is a promising approach
in the battle to boost America's manufacturing prowess. Here's what's required
to make the process work:
Worker cooperation A no-layoff pledge is necessary to assure workers their
suggestions for productivity improvements won't cost them their jobs.
Management commitment Top management needs to back the kaizen team with the
resources and authority to make sweeping changes, if need be, in search of
radical efficiency gains.
Measurable goals and results Besides validating the team's successes,
measuring enforces a reality check.
Diverse teams Kaizen teams should draw from disparate backgrounds--not just
manufacturing--to elicit a variety of ideas.
Bias for action The teams should favor immediate action on many fronts,
opting for small improvements over costly, technological big fixes.
Follow-up Not everything can be done in a week, so follow-up work is needed
to implement medium-term suggestions.
James B. Treece