Staying Power BorgWarner unit in Muncie looks to the future as it celebrates its 100th anniversary.
John H. Sheridan
BorgWarner, Diversified Transmission Products Inc., Muncie Ind.
At a glance
A pioneer in adapting high-speed flexible machining centers to aluminum and magnesium cases, the Muncie plant partnered with Ingersoll International Inc. to develop the state-of-the-art Ingersoll 600 Series machining center.
Workforce of "owner-operators" is 100% multiskilled.
100% on-time delivery.
Reduced warranty costs by 89% in five years.
Received ISO 14000 certification this year.
Won 1999 Indiana Quality Improvement Award. Long-time residents of Muncie, Ind., still refer to the operation as Warner Gear -- the company founded in 1901 by Henry Warner to manufacture differentials for Ransom E. Olds' earliest automobiles. For a full century, what is now BorgWarner Inc.'s Diversified Transmission Products unit has demonstrated a knack for survival, recording its share of business triumphs and surmounting many challenges along the way. "Over the years," observes Wayne Wilson, a plant operator and vice president of UAW Local 287, which represents the workforce, "we've made changes when we needed to make them to stay competitive -- changes in our products and in our production processes." Shortly after World War II, for example, BorgWarner engineers in Muncie developed the Ford-o-Matic transmission, the first automatic transmission ever installed in Ford Motor Co. vehicles. The manufacture of automatic transmissions became a mainstay of the Muncie operation until the Big Three automakers decided to pull production of automatics in-house. In response, the BorgWarner plant on Kilgore Ave., which dates back to 1928, shifted gears, so to speak, and became a major supplier of manual transmissions as well as transfer cases -- the complex gear-laden devices that control distribution of power in four-wheel-drive vehicles. With demand for manual transmissions waning, the company reached another critical juncture in 1996. "A tough decision was made that to save the business as a whole we had to exit from the manual transmission business, with the vision that one day we could grow our transfer-case business enough to fill up the vacated floor space and return the [laid-off] employees back to work," explains Ron McCoy, vice president, operations, who heads the management team at Muncie. The last manual transmission rolled off the assembly line in 1998. To fulfill its new vision, the plant stepped up its focus on quality and productivity, rolling out a multifaceted "Visual Factory" initiative that, among other things, has prompted the relocation of 558 machines in the last two years to improve workflow, reduce WIP inventory, and enhance quality. As hoped, transfer-case production has increased steadily, but the plant's narrower product focus had one major drawback: It found itself heavily reliant on a single customer. About 90% of this year's output of 950,000 units is destined for Ford vehicles. The other 10% will go to Korea's Daewoo Motor Co. Ltd. However, early next year the customer base will become more diversified. In a major coup, the Muncie team won a five-year contract that will ramp up to production of 350,000 transfer cases a year for General Motors Corp. (Significantly, GM holds a one-third stake in the plant's major competitor -- the East Syracuse, N.Y., plant of New Venture Gear Inc.) As the folks in Muncie put the final touches on plans for their 100th anniversary celebration this month, extensive retooling and other preparations were well under way to accommodate the new GM business that will indeed occupy much of the floor space vacated in 1998 and create several hundred new jobs. The 1,300-employee plant's stellar quality performance, including an extremely low (37 ppm) customer reject rate, clearly attracted GM's attention. "You've got to believe that GM came here for a reason -- taking business from their own joint venture, so to speak," says Randy Heider, director of manufacturing for small transfer cases. "I think a lot of it had to do with the quality numbers we presented them. You could see the twinkle in their eyes." Thorough testing of finished products is one reason for the 37-ppm reject rate. Extensive use of electronic sensors and other error-proofing devices also plays a major role, along with design for manufacturability (DFM) practices and a concerted effort to deploy statistical process control (SPC) techniques effectively. Over the last five years, the plant has slashed customer rejects by 97% while reducing internal defects by 70%. Equally important is a long-standing culture of employee dedication to quality. "When we were hired in, quality was driven into our heads, not by the management people, but by the people who worked here who are now retired," says Michael Brown, a 30-year plant veteran and treasurer of UAW Local 287. "They let us know that they were proud of making the best-quality transmissions in the world, which is what we were making back then. It didn't take long for us to develop the same attitude." That culture is continuously reinforced, not only during weekly quality meetings but also with periodic bus trips in which groups of 40 or more employees visit a customer assembly plant. "We want our people to talk to the assembly operators who receive their products," says Dan Mills, director of quality and manufacturing engineering. When McCoy took the helm at Muncie two and a half years ago, he intensified the push for quality and introduced a "Listening to Processes" theme aimed at reducing variation. McCoy has emphasized an important lesson that quality guru W. Edwards Deming impressed upon him nearly 20 years ago -- that there is a right way and a wrong way to implement SPC. Applied correctly, SPC can guide the reduction in process variation and target performance "so that you get it nailed exactly to what the customer needs," McCoy says. It is not a matter of creating a multitude of SPC charts, he stresses, but of ensuring that the characteristics you're charting are the right ones. "You have to understand which variables you need to control to truly reduce variation," he explains. "It is painstaking hard work. It requires design of experiments, looking at your critical process characteristics, and a lot of trial and error." Peering out from a glass wall near the entrance to McCoy's office is a stuffed replica of Garfield the Cat, the adopted symbol for another major plant initiative: cost attack teams, or CATs. Garfield is an appropriate mascot since his creator, cartoonist Jim Davis, works out of a studio in Muncie. Conceived as a way to increase employee empowerment, CATs exist at two levels. Any employee can spontaneously assemble a CAT team to implement an idea that will improve quality, productivity, or safety. "It is not a suggestion program," says McCoy. "It's an action program." In one instance, machine operator Eric Glaze came up with a poka yoke device -- attaching a bolt to the fixture that holds a particular casting in a high-speed machining center. The bolt ensures insertion of the proper casting (which looks nearly identical to another casting). In the past, insertion of the wrong casting caused destruction of the machining spindle about 10 times a year, incurring a $25,000 repair cost each time. With the help of a toolmaker, Glaze installed the error-proofing device in half a day. A more structured version of the initiative, dubbed Fat CATs, involves teams chartered by management to tackle larger projects that may require extensive layout changes or significant capital expenditures. A number of Fat CAT teams have been launched under the Visual Factory initiative to introduce automation or to implement one-piece-flow concepts in component manufacturing cells. Over the last two years, the combined efforts of CATs and Fat CATs have yielded approximately $6.1 million in annual cost savings and productivity gains. No wonder plant operator Doug Chalfant asserts: "We know what it takes to stay in business. And we hope we'll be around to celebrate our 150th or 200th anniversary." Web-Exclusive Best Practices By John H. Sheridan Benchmarking contact: Ron McCoy, vice president, operations, [email protected], 765/286-6600. Weekly SPC audits To ensure that statistical process control (SPC) methods are deployed properly, the Muncie plant has a 10-person team of process engineers and hourly SPC facilitators who instruct others to implement SPC in a way that will reduce variation in critical performance characteristics. At each process step where SPC charts are created, an instruction book includes a "Why" statement, detailing why the metric is important to ultimate product performance. In addition, every SPC chart in the plant is audited on a weekly basis by a member of the management team to ascertain: (1) whether operators are following instructions properly and (2) when a process change is made, whether the new procedure has been standardized to permanently "capture" that improvement. "That's how you get to Six Sigma quality levels," says Ron McCoy, vice president of operations. "You come up with the best way to do something, and everybody on all the shifts does it that way. It is the Toyota Production System concept of standardized work." "SPC gives you the data to support making changes," adds Lowell Drill, director of manufacturing for large transfer cases. "When operators see that you are willing to react to the data -- perhaps by spending money to rebuild a piece of equipment that might be worn out -- then they can see the benefit of it." Design Bursts Survival as an automotive industry supplier requires almost incessant attention to cost reduction these days. At Borg Warner's Muncie plant, "design bursts" help to achieve that. These are two-day brainstorming sessions aimed at reducing costs related to new products. Participants in the sessions have included professors from nearby Ball State University, purchasing staffers, members of Borg Warner's design engineering team in Sterling Heights, Mich., and employees from other company divisions that impact the targeted product in some way. The mix of participants also may include people who are not very familiar with the product -- but whose questions often trigger new insights. "The idea," explains Dan Mills, director of quality and manufacturing engineering, "is to find ways to take as much as 20% of the cost out of the product. Our latest design burst generated a four-page list of nearly 100 ideas. Some of the ideas we get are pretty wild. But if you get 10 good ones, it is well worth the time." Constraint Management Although the plant has been emphasizing one-piece -- flow-and has achieved it in some of its component production cells -- a number of cells maintain buffer stocks between certain operations. It is an element of constraint management, which the Muncie team learned from consultant Eli Goldratt in the late 1980s. "Buffers happen for lots of reasons, but what you try to do is buffer the constraint so that it is not ever sitting idle," explains Randy Heider, director of manufacturing for small transfer cases. "If that constraint sits idle, you've lost that piece -- or that [production] time." Every manufacturing cycle has a constraint, he adds, "but it is different in every cell." For example, in a ring gear cell, the first operation is broaching, which cuts the inside of the gear teeth. While the broach is relatively fast, it feeds a bank of perhaps three slower machines called shapers, which cut the exterior of the gear. Predetermined buffers are maintained in front of the shapers. When the buffers are full, production at the preceding operations stops. An ongoing mission of the manufacturing team is to try to reduce the size of the buffers as the production process is improved. One result: a 38% decrease in WIP inventory over the last five years. Workforce Flexibility Adoption of manufacturing cells in the mid-1980s generated significant productivity gains at the Borg Warner plant. For example, the cycle time to produce a main shaft was compressed from three weeks to an hour and a half. However, a quantum leap in worker flexibility was necessary. "Prior to 1983, we had 37 different direct-worker occupations, such as 'blanker' or 'grinder' or 'heat treater,'" recalls Randy Heider. "But in our 1983 union contract, we changed to a single category -- 'owner-operator' -- which allowed us to convert to U-shaped cells. The operators got a lot smarter because they had to learn to run anywhere from two to five different kinds of machines." The plant's most recent union contract turned staffing of overtime assignments over to hourly workers. Each shift in the plant now has a union-appointed "overtime/movement coordinator" who is charged with ensuring that the opportunity for overtime work is offered on an equitable basis. This frees up supervisory personnel for other tasks. "In the case of overtime, we decide how many people we want and what cells or assembly lines we want to run," Heider notes. "Then we give that number to the hourly person and he gets the people. . . . Has it been perfect? Absolutely not. But I don't spend much time arguing about overtime anymore." In-Line Induction Heat-Treating One technological advance that has improved workflow through the Muncie plant has been the installation of induction heat-treating equipment in some production cells. For example, in one reduction-hub cell that feeds the final assembly line, induction heat-treating takes just 30 seconds. An intense magnetic field causes the treated gear to quickly glow a bright orange color. A sensor monitors the temperature of the part during heating and will reject it if tolerance limits are exceeded. "Induction heat treating has been a real boon to us," says area manager Mark Bell. "It's helped us to streamline our flow. We can do one part at a time rather than have to float all the inventory we'd need [for a conventional heat-treat oven]. In the past, we had to have a two-day bank of parts to flow to heat-treating." Also enhancing workflow are the small coordinate measuring machines (CMMs) deployed in high-velocity casting-machining cells. The "Discovery" units, much smaller than conventional CMMs, take nearly 100 measurements on a machined surface in six minutes. Not only do they gage the dimensions of the workpiece, they also check the temperature and automatically compensate for shrinkage during subsequent cooling. "In the past, when we sent a piece to the [large] CMM for measurement, we had to wait about 30 minutes for the part to cool down," explains supervisor Shane Sherrell. "And while you were waiting, you could be running out-of-tolerance scrap." CMM measurements are taken on the first four units out of the machining center after a changeover -- and also after a shift change.