The story behind GM's costly J-car launch

LORDSTOWN, OH -- it looked great on paper: An ingenious plan to convert General Motors Corp.'s huge assembly and stamping complex near this small northeastern Ohio town into a super-efficient manufacturer of high-quality, low-cost small cars -- the new-generation 1995 J-body Chevrolet Cavalier and Pontiac Sunfire.

The original blueprint called for Lordstown to be the cornerstone of a $1 billion investment involving three plants with total annual capacity exceeding 550,000 and 3 million over the product's life-cycle. Lordstown capacity alone, based on the 3-crew/2-shift manning put in place in 1992, is 400,000 cars per year.

The basic plan hasn't been tossed in the waste basket, but it's badly crumpled after a painfully slow launch that began here last Aug. 8. Now, nearly eight months later, Lordstown still is struggling at well below targeted capacity, which is said to be 1,600 cars a day. In February the plant built just 677 cars daily.

What went wrong at Lordstown? GM executives are nebulous to a fault when asked that question. But they generally agree their reach exceeded their grasp, and tried to put a bright spin on the situation by insisting that issues resolved here eventually will go a long way toward slashing bloated manufacturing costs at other GM plants. WAW interviewed key GM executives, automotive analysts, union officials and numerous others, seeking reasons behind the lethargic Lordstown ramp-up. The all-new stamping operation has given GM planners headaches. And there are numerous reports that some components, systems and processes haven't meshed as planned. One engineer, for example, says the window-cranking mechanisms didn't work as designed, and the line had to be stopped.

There's not much tolerance in Lordstown's totally new lean-production system for mistakes. Patterned after Toyota Motor Corp.'s vaunted system, which also is used at the GM-Toyota New United Motor Manufacturing Inc. (NUMMI) plant in Fremont, CA, the Lordstown scheme represents a sea-change in the way GM does things. Even so, several key players at Lordstown have worked at NUMMI and brought that experience to the changeover. (A NUMMI executive notes, however, that he offered to share ideas with the folks at Lordstown but found a cool reception).

A major element of Lordstown's new processes is Toyota's andon system that detects problems and deals with them immediately. Signboards throughout the plant show whether production is running normally or not, effectively monitoring work flow at the job-site. If there's a problem detected in a malfunctioning machine or by a worker who pulls a cord, the sign lights up and music or chimes play on a loudspeaker. If the problem can't be solved quickly, the line actually may shut down.

Despite some 800,000 hours of worker training, including familiarity with andon, sources tell WAW that adapting to this new system alone has created time-consuming problems that have crimped output. Regardless of the causes, Lordstown's sluggish launch has hurt. The market was hot when the cars were supposed to arrive in sufficient volume. It since has cooled, making it tougher to recapture lost ground. Meanwhile, it's creating a mammoth cash drain. A highly regarded analyst says it's costing GM another $1 billion through the double whammy of lost revenues and stiff ongoing costs. GM officials won't confirm that number, and in any case say it's not relevant; getting the plant to build high-quality cars is the mission.

Still, Lordstown's malaise is taking its toll. Cavalier in recent years has been Gm's best-selling model and Chevy's price leader. It was the nation's fourth-highest seller in 1993. But due largely to Lordstown's minuscule '95-model output during the final five months of 1994 -- only 21,000 '95 J-cars left the plant -- Cavalier sales last year plunged 31.6% to 187,300 from 274,000 in 1993. Combined Sunbird/Sunfire sales (the Sunbird name was dropped in the '95 changeover) also sank by nearly a third to 68,000 from 93,000.

Automakers typically lose sales and market share during major tear-ups, but the shift from the old models to the new took just four weeks, an amazingly short time for a project of this magnitude. The first cars reached dealers in October, but since then they've been as rare on the nation's highways as folks heeding the 55-mph speed limit.

Lordstown output has climbed steadily since launch, reaching nearly 19,000 in February, but dealers still don't have enough stock to mount a meaningful sales effort -- and Chevy and Pontiac have yet to launch significant marketing campaigns.

Chevy has booked more than 116,000 orders for Cavalier, and Pontiac dealers have ordered 73,000 Sunfires. At February's production pace, it would take 10 months simply to fill that pipeline.

A further measure of Gm's dilemma: During the first five months of model-year 1995 (October to February) Cavalier sales ran 72% behind the prior year: 32,000 vs. 115,000. Sunbird/Sunfire sales were off almost 70% at 11,800 vs. 41,000.

Lordstown was scheduled to build an estimated 27,000 cars in March, and April output reportedly will match last year's 34,000 output. "We'll be in stride during April and make bigger strides in May and June," promises G. Richard Wagoner, president of Gm's North American Operations (NAO). "I'd love to have had them sooner, but we were trying to drive the profit and quality issues."

Lordstown aside, the other two pieces of the J-car pie apparently are falling smoothly into place. The Ramos Arizpe plant in Mexico, which uses stampings supplied by Lordstown and has basically the same suppliers, also launched J-car assembly last August. Compared to Lordstown, Ramos is a low-technology operation that made the changeover to the new model with relative ease. It has been running at a 10,000-per-month clip since January, and actually is operating well above the scheduled 70,000 to 80,000 annual rate. Ramos J-cars are sold in Mexico and exported to Canada plus smaller markets.

The third J-car site is Lansing, MI, home of Gm's Lansing Automotive Div. (LAD), which is responsible for all three J-car plants. Lansing began making Cavalier coupes on March 21 on the same line as its N-body cars (Grand Am, Achieva and Skylark), demonstrating a new dimension in assembly flexibility as GM gears up to build the new successors to the N-cars in 1997. Lansing now picks up some of Ramos' production under a GM/UAW special agreement. Convertible versions of Cavalier and Sunfire will move into production later this spring at a nearby off-site facility. Total Lansing J-car output, including ragtops, is scheduled at between 70,000 and 100,000 yearly.

As these plans suggest, gearing up for the new-generation J-cars -- their first major overhaul since they debuted in 1981--has not been a small undertaking.

The key to it all, of course, is the giant 3.7-million-sq.-ft. (344,000-sq.-m) Lordstown complex, a facility around which controversy long has swirled (see sidebar p. 27) but in recent years has been the focal point of remarkable cooperation between management and the United Auto Workers union -- at least until now. There's increasing grumbling in UAW ranks that GM may be violating contract provisions relating to sourcing and staffing.

The problems are all the more damning because the '95 J-cars originally were scheduled to arrive in 1992 but were twice postponed, giving GM planners plenty of time to check out all of the angles.

"The big issue," Mr. Wagoner admits, "is we bit off a big chunk. There were three orders of magnitude -- three buckets. If it had been the same plant, same processes and same productivity it would have been easy. But if you vary all of those things it becomes more complex.

"And at Lordstown, we had an all-new product, significantly new concepts on processes to improve quality, and much higher productivity targets (30%). We were challenged to build small cars at a profit. We knew what we were doing, but I guess you can say we were overly optimistic regarding bringing it up to (full) production."

GM officials quickly put down speculation about possible causes for problems. Equipment and component sourcing, they say, was completed before controversial purchasing chief J. Ignacio Lopez de Arriortua arrived in the U.S. three years ago and put the heat on suppliers to cut their costs. It hasn't been "payback time," as some suggest, says Mr. Wagoner.

Suppliers also are basically exonerated. Mr. Wagoner says he'd be "hard-pressed" to point to any serious supplier flaws, "given what we set out to do," and adds: "I'm not out huntin' suppliers on this issue."

He also dismisses reports that Lordstown's problems can be traced to Gm's financial woes during the build-up period. Once set in motion, GM hasn't stinted on the J-car investment plan, he maintains.

James R. Wiemels, who replaced Thomas J. Davis as GM vice president and LAD general manager last Dec. 1, traces the slow start to myriad issues coming together. "It's not just one thing," he says. "It's a compounding factor."

A former high-ranking GM executive engineer and student of its organizational chart lays Lordstown's problems squarely on "the leadership issue." Although Mr. Wiemels' assignment to replace Mr. Davis (now vice president-engineering at the NAO Truck Group) was not publicly, at least, tied to Lordstown's problems, this source has no such qualms. "If Tom Davis had said `We're not ready,' it would have reflected on him." Instead, the project moved ahead and has become a giant fix-up, he adds.

One big issue relates to reducing the workforce by 2,400 through attrition and re-assignment, then retraining and reshaping the roles of the 8,800 who remain at the assembly and adjacent fabrication plant. "When you take that many people out of the operation, not only does everyone's job change because of the new product but you (also) have a lot of movement of people," says Mr. Wiemels.

The overall goal was to reduce the man hours it takes to build a car from a costly 32 to the low 20s -- a sizable 30% leap in productivity and the chief determinant to quality and profitability.

A constant theme voiced by GM is that quality won't be sacrificed at Lordstown to the altar of faster output. Yet in February GM recalled 34,000 Sunfires and Cavaliers for a missing torsion-bar weld that can cause steering problems. WAW sources say the parts are made in Lansing by LAD. That amounts to practically the entire first six-month run of new J cars.

Quality of an early J-car WAW tested also was far from top-notch when it comes to exterior fits, and two Sunfires were spotted in SAE Congress displays early last month with dubious fit and finish. Still, the new models have picked up praise from several publications, and Mr. Wagoner maintains that "in the aggregate, on a comprehensive scale we're fully meeting our quality targets."

GM won't divulge Lordstown's original ramp-up schedule, but one expert, John Russo, director of labor studies at Youngstown State University, estimates it was 1,600 cars a day. Sources say that plan was re-written in December to 1,000 cars a day. By mid-February Mr. Russo, who has students working at Lordstown, says output was 65% of the revised goal.

A GM engineer confirms the 1,000-car-a-day bogie. "We're still trying to hit 50 an hour or 1,000 a day," he says. "We've hit that rate a few times, but right now we're in the 40s somewhere. Basically we're at the point where we're running fairly well."

Meantime, dealers are playing a waiting game. A Detroit area Pontiac dealer said recently she had four Sunfires in stock, compared to the 40 Sunbirds on the lot a year earlier. "We sell every one we get in," she says.

The scope of the Lordstown makeover has been mind-boggling. It included moving more than 100 presses in the stamping plant, taking out 20 presses and putting in extra die carts and automation. Other maneuvers include putting some sub-assemblies in the fab plant, allowing Lordstown to finish the 94s and begin conversion of both the fabrication and assembly facilities so they'd be ready for the '95s.

The plant's old side-gate body assembly system using 55 tools to build up bodies was replaced by a state-of-the-art robogate system using just one tool, with the expectation of producing more consistent bodies. In total, the plant went from 50 robots in 1993 to some 200 today. Also, the plant's entire conveyor system is new.

"When we changed the conveyors we changed some of the flooring, too," says one engineer, who has spent a great deal of time at the plant. Bodies reportedly were streaming out of the stamping plant while concrete was still being poured in the assembly plant to replace sections of the wooden block flooring.

While that was going on, 450 of the 2,400 workers cut from the assembly plant were transferred to the stamping plant, which was shifted to a three-crew/two-shift operation synchronous with assembly. Another 350 were loaned to Lear Seating Corp., which supplies seats on a just-in-time basis from a nearby facility. About 300 were transferred to other facilities, and the balance took retirement packages. With all of this shifting, it's little wonder things haven't meshed as visualized.

Regardless of whether the Lordstown changeover was too much to swallow in one gulp, Mr. Wiemels says that "The window of opportunity was there -- a chance to bring in a brand-new product, change so much of the workforce at one time and to teach them the Toyota quality system. The complexity of that has brought us to this very methodical learning curve."

Mr. Wagoner credits LAD for doing as well as it has so far, and he comes close to agreeing that GM goofed in not spending more time up front. "We've got to do a better job of releasing dies and tooling earlier," he confides.

Another highly placed GM executive says: "We underestimated the complexity of re-doing the entire body area. They should have started a year earlier."

"Oh yes, they severely underestimated the changeover," says Robert Chambers, president of UAW Local 1714, which represents stamping plant workers." They got off on the wrong foot with all the automation and other stuff they put in."

Despite Mr. Chambers' assertion that his membership is "shipping by and large the best metal we've shipped in the history of the location," many point to snafus in the Lordstown stamping operation as one of the biggest reasons for the slow start.

One explanation relates to re-engineering the stamped parts. Chevrolet, Pontiac and the GM Design Center worked with LAD to design panels that could be processed with fewer die hits. At Lordstown, which stamps 90% of the sheet metal for all J-car production as well as for Saturn, die hits are reduced from 5 to 3.1 per panel, with many components now stamped together simultaneously.

Fewer die hits equals fewer tools and faster manufacturing, which means less cost. But with all-new processes, sources say there have been problems such as fenders with stretch marks and pieces that don't fit as they should. Analysts and even a high-ranking GM executive suggest that if dies and equipment arrived at the plant sooner, there would have been more time to de-bug the system.

One analyst who spent time in the facility says equipment scheduled to go in the plant in July was still going into place in October. Dies also were late, he adds. Says a high-ranking GM executive not directly involved with Lordstown: "The (tooling) is coming in from all over the world, but you have to remember that productivity improvements were built into the plan. From the beginning we didn't want to lose that. We could have thrown a bunch of people at it, but we didn't want to go backwards." He is clearly referring to Gm's old ways of problem solving.

Mr. Wiemels says reports that stamping tools arrived late are "absolutely false," adding that "all systems are up and running." He allows, however, that "we are still ironing out issues, tweaking the equipment and continuing to make progress every month as the numbers indicate. We are very pleased with the quality and we're making progress every week."

"If there's a silver lining, in the end they should save $500 or more for each body," says David E. Cole, director of U-M's Institute for the Study of Automotive Transportation. He says it'll pay off in higher productivity throughout the GM empire.

Some of the few outsiders allowed into the plant observe "filled" repair areas, which Frank Martin says is traditional after any changeover. Mr. Martin is financial secretary of UAW Local 11 12, which represents assembly plant workers.

Others report interior trim quality problems that have produced delays. "Supplier problems probably have hurt us somewhat," says one engineer close to the seene. "But you don't need many to cause a real problem: if you close down one area for five minutes, you lose six cars."

Even Mr. Wiemels apparently agrees that the main "issue" in both the stamping and assembly operations is getting the technology, specifically automation, to work right. "We're still ironing out issues in the body shop -- issues like technology," he says.

Mr. Russo, the Youngstown State professor, observes that "GM historically has a tough time getting the technology and the manning right. They tend to err too much on the technological side." A shining example: GM's Poletown plant in Detroit, a technological marvel beset for years by shoddy quality and production quirks, yet no discemible increase in productivity.

The UAW's Mr. Martin also understandably blames the robots. "It's the automation," he says. "When one flute is out of tune, the whole orchestra sounds bad."

There may be some substance to that. Mr. Wiemels underscores that the smooth J-car changeover at Ramos Arizpe and Lansing clearly demonstrates the manufacturability designed into the new J-car. But it's a different story at high-tech, high-volume Lordstown.

A source in GM's robotics department says the Mexican plant has only about 30 robots, yet it ramped up from last August through November faster than Lordstown. Mr. Wiemels says "less complexity" is the reason the startups in Mexico and Lansing have gone so well. In Lansing, he says, "We're not taking any down time to start up the J-car flex operation.

"This car is definitely easier to assemble than most any car we've done in the past because of all the (design for manufacturability) things we've done up front," Mr. Wiemels explains. "The fact that Ramos came up to speed so quickly and that we're able to flex it in Lansing proves out that the vehicle builds well and builds easily, and that extrapolates into improved quality."

One result of Lordstown's slow pace reportedly has caused tempers to flare between engineers at LAD headquarters and at the Ohio plant. Observers suggest Lordstown engineers have gotten more Lansing "assistance" than they need, which is perhaps understandable given the heat LAD has been under. Mr. Wiemels admits to "rare" spats between the two groups.

Another issue facing the plant is employee relations. Mr. Wiemels says when you're trying to change the work culture and system to set the stage for the future, it takes time.

"When you involve the people, you get more than all the cooperation you need," he says. "People are taking on a lot of new roles that are different than they've had in the past. And a lot of new responsibility. If you do the proper education up-front, people are more than willing and cooperative. People have not been an issue with this startup."

Yet the finger of blame still is being pointed around internally, as all involved look to find explanations for Lordstown's woes. A memo from UAW Local 1112 Shop Chairman Al Alli to Cal Rapsom, vice president and director of the union's GM department, obtained by WAW shows a somewhat less than totally cooperative attitude by the union.

"Management continues to smokescreen a line-speed decrease to gain manpower," memo reads. "Management states that this is only part of their model start-up and acceleration plans. They mouth this line of propaganda while saying we better watch out or Lansing will get our work. Management has forgotten that they have caused the start-up problems and forgotten the commitments of job security.

"Management's model start-up problems have led them to say they side-step our Local Memorandum of Understanding on Overtime regarding overtime allowed past ten (10) hours," Mr. Alli's memo continues. "Management's side-stepping position is that the Understanding doesn't apply, since we will be in a model acceleration the whole year. Again, Management wants membership to pay the price for their mistakes."

GM and analysts alike suggest the delays at Lordstown, while costly and perhaps embarrassing, will become less significant when weighed over the new J-car's life-cycle - if the quality's good.

There's an even more pressing reason to get the quality down pat: Starting this fall, GM will begin shipping 20,000 right-hand-drive J cars built here to Japan to be sold as Toyota Cavaliers.

Whatever Lordstown's launch is costing GM, the lessons learned may be invaluable, says U-M's Mr. Cole. "It may be cheap compared to what it would cost long-term if they didn't make these changes," he states. "It may be a big hit (on profits) over the last year, but if you're going to be an affordable car company you've got to have the technology. When they get it working, they'll have a big competitive advantage."

The snail's-pace Lumina/Monte Carlo start-up a year ago at Oshawa, Ont., gave GM fits as well, but the pain was worth it, says Mr. Wagoner. "We're getting 25% higher productivity in Oshawa and our cost improvements are actually better than that," he says. "The quality is dramatically higher and we're getting 66 per hour. Today it's a winner. We had a lot of bruises, but it's running great. It shows the upside of the corporation."

In that sense, he'd welcome an Oshawa replay at Lordstown.

The Rock and Roll Hall of Fame is in Cleveland. So there's an elegant irony in the fact that General Motors Corp.'s Lordstown, OH, plant pops out of nowhere in particular along the Ohio Turnpike not many miles east.

Both have working-class roots, a history of rebellion and a certain level of discomfort dealing with middle age.

The music speaks for itself, but GM veterans remain puzzled over how a relatively prosaic auto assembly complex became a symbol of the '60s youth rebellion, the '70s decline of the U.S. auto industry, battles for the soul of the United Auto Workers union and now the focal point of GM's rocky drive to reach world-class lean production.

Although Lordstown made headlines worldwide more than 20 years ago, it may be the only auto plant in the world to twice make the cover of Newsweek.

"It really is just another auto plant," a GM insider mutters.

By auto industry standards, Lordstown has barely reached middle age. It opened in 1966, producing the Chevy Malibu Sports Coupe, but didn't reach the national consciousness until the '70s.

At that point, GM made what may be one of the biggest public relations blunders in U.S. auto history. The plant was given the Chevrolet Vega, aimed at the exploding youth market. Cool idea. Young people making cars for their peers.

The problem was the workforce shared a lot in common with young graduates of the '60s. They hated traditional authority, used drugs, and had delayed starting families. As one result, showing up for work was not always a top priority.

In what became a legendary quote, a young Lordstown worker was asked by a top UAW official why he was only showing up four days a week. "Because I can't make enough money in three," was the thoughtful response.

Seizing on all the national attention, local Lordstown UAW officials began openly taunting union leadership for being out of touch with a new generation of workers. In the end, both the ho-hum, problem-plagued Vega and Lordstown's Age of Aquarius workforce vanished. So did the GM Assembly Div. (GMAD), notorious for its bare-knuckles approach to labor relations, which ran Lordstown.

Since those tumultuous early days, the Lordstown workforce has largely become part of solutions rather than problems. Workers there have taken the lead in agreeing to try new solutions to manufacturing and productivity problems.

But the starkest testimony to the fact times have changed at Lordstown is that the plant still exists given that local worker attitudes were an important factor in deciding which GM plants would survive the spate of closings in recent years.

On the other hand, is it just possible that the young guy who in 1971 wore a "Bleep Authority" T-shirt on the assembly line is now a grandfather?