Plant operators frequently play the role of garbage collectors at the expense of manufacturers. Instead of producing, they're gathering scrapped parts and materials to be recycled, reused or discarded. Facing foreign competitors who already are able to produce parts and materials at a significantly lower cost, U.S. manufacturers cannot afford the additional labor costs incurred from handling excess scrap.
Although manufacturers expect some scrap will be generated during the production process, truly lean operations strive for little or no scrap. These companies employ a variety of continuous-improvement methodologies, including Six Sigma, Multivariable Testing (MVT) and simple chart analyses, to reach their scrap-reduction goals.
Manufacturers tend to be more successful in their scrap-reduction efforts when they use some type of scientific approach rather than basing their plans of action on experience, says Ralph Rio, research advisor, ARC Advisory Group, Dedham, Mass. "Unfortunately, most manufacturers are going by tribal knowledge, meaning someone has an opinion about how to fix a particular operation or to reduce scrap and, by dominance of their personality, has a change implemented," he says. "Sometimes it works, but usually it doesn't."
Instead, Rio says manufacturers should depend on fact-based knowledge, often derived from information entered into databases by plant-floor operators that is then represented on a simple Pareto chart -- a bar graph with values plotted in descending order starting from the left-hand side.
The quick and simple approach to solving scrap problems is what David Cochran, vice president of operations for QualPro Inc., an MVT training and consulting firm based in Knoxville, Tenn., stresses to his clients. MVT involves a brainstorming process in which several people with different roles throughout the company and plant operations list ideas to address inefficiencies. Manufacturers then narrow down the list based on which ideas are the most practical to implement, usually depending on how costly or time consuming the project will be, says Cochran, who co-authored with Charles Holland "Breakthrough Business Results with MVT."
"If something is not easy to test and implement, we set that aside because what we found is that these easy and quick and inexpensive changes very often can create these big breakthrough results," he says. The next step with MVT is to test the ideas in live situations. The ones that produce positive results are implemented while the ones that fail are dropped.
Using these tools to analyze processes at the operations level is a common way to measure and identify scrap problems, but material waste also can be prevented by implementing changes at the engineering stage. In the auto industry the first line of attack against scrap often is taken at the engineering stage, says Ron Krupitzer, vice president of automotive applications for the American Iron and Steel Institute.
For instance, stamping plants sometimes use steel blanks that are larger than needed to ensure there's enough workable material during the tryout process. Over time, auto manufacturers have found ways to fine-tune the engineering process so they can use the smallest-size blank possible. One way Tier One auto parts suppliers have trimmed blank sizes is by using more computer modeling applications to predict and simulate the stamping operation, Krupitzer says. This gives them a better idea of what steel grade will produce the least amount of scrap.
Auto manufacturers also have experimented with stronger steel to reduce scrap, sometimes in the assembly of inner-door panel hinges where the need for additional reinforcement parts results in more waste. Now automakers can use laser-welded blanks, or blanks made from higher-strength steel to eliminate the need for hinge supports, Krupitzer says. Higher-grade steel may also be used in the stamping process to prevent splitting and lower defect rates.
Focus on Obvious Opportunity
For the Hexacomb division of Pregis Corp., an $850 million protective packaging company based in Lake Forest, Ill., the scrap problem was paper and glue used to create die-cut packaging. When Hexacomb President Bill McBee joined the division in 2003, he was looking for ways to reduce costs. He started by focusing on the most obvious cost-cutting opportunity: waste. During the process of converting paper into die-cut packaging, Hexacomb was scrapping the equivalent of approximately one out of every four rolls of paper it purchased, according to McBee. He estimated the cost of wasted paper, glue and labor at $4.5 million.
McBee started his scrap-reduction initiative by selecting Hexacomb's top-performing plant in Trenton, Ill., for an MVT experiment. A brainstorming session with all the plant's hourly and salaried employees resulted in 350 to 400 ideas. McBee says it's critical to involve the people who are most familiar with the plant-floor processes in these sessions. "You start off by shooting all the engineers -- they don't really have the answers," jokes McBee. "Then you get all the operators and material handlers and everyone you can who really understands what's going on in the process . . . and you go through four days of SPC (statistical process control) training, using statistics from the plant that you're going to work with."
From there, the plant staff narrowed the 350 or so ideas down to 24 variables to test by eliminating duplicate suggestions and less-practical ideas. Over a 48-hour period, plant-floor operators conducted the tests and recorded the results. The plant found three factors that reduced scrap, three that increased waste and 15 that had no effect at all, McBee says.
The changes that worked -- replacing old glue rollers with new ones, setting scheduled speeds for the machines that make the cores and tweaking the speed of the saws that cut the paper -- reduced scrap at the Trenton facility by half.
Hexacomb has extended MVT to other plants and achieved further scrap savings. In 2005, the company conducted an MVT of 17 machines at eight different plants, but McBee cautions that the method is more effective when it's narrowly focused. "MVT works best if you have a situation where you focus on one process -- for example, one machine, one plant," he says. "When you try to combine all of them, you get a lot of averaging, so you don't get nearly the impact when you do a whole bunch of them with the same factors on the same days."
About six years ago, Cummins Inc., a Columbus, Ind.-based manufacturer of large diesel engines, began making Six Sigma part of the company culture. One of Cummins' goals with Six Sigma was to reduce the amount of scrap at its fuel systems plant. Through the Six-Sigma process of defining, measuring and analyzing, the company found several opportunities to reduce material waste. One of the key people in implementing these improvement projects was account manager Ginger Lirette. In 2005, Lirette received one of the company's first J. Irwin Miller awards -- named after the company's former CEO -- for completing 13 quality improvement projects that saved the company nearly $14 million.
One project headed by Lirette was to analyze the gauging on the fuel systems injector line where the company was experiencing a high defect rate. Using the Six Sigma methodology, Lirette discovered that the gauging was not accurately simulating an operating engine. "Based on that study, we were able to rework our gauge to be more real-world like, and we were able to save quite a bit of time, scrap and labor," Lirette says.
Another Six-Sigma project the company undertook revealed that one of the operations intended to improve the roundness of injector plungers actually was causing defects. By removing this step from the process, the company was able to improve the quality of the plungers and reduce scrap. "How often does that happen?" Lirette asks, then answers: "[It's] probably not very often that you actually get to take an operation out, so you're reducing your inventory, you're reducing your labor time and getting a quality improvement."
In 1999, General Cable Corp.'s 56,000-square-foot Moose Jaw, Saskatchewan, Canada, plant was operating at approximately 4% gross waste, according to plant manager Ray Funke. At the time, the factory, which was a winner of IndustryWeek's 2005 Best Plants award program, was using some rudimentary tools to measure scrap, but it wasn't providing the detail that was needed to produce significant results. So the plant developed detailed metrics of every cause of scrap from every workstation in the plant using several lean tools.
The plant utilized a Pareto-analysis chart to engage its workforce in its scrap-reduction efforts by showing them where scrap was occurring. Funke says getting production workers involved in the scrap-reduction process was a key turning point for the plant. "It was a major initiative because our company in 1999 wasn't in a very good position from a financial standpoint with the market downturn in wire and cable, and we were obviously trying to cut costs in any way possible, and scrap was clearly a big opportunity for every plant within our company," Funke explains.
By using the charts and involving employees in the process, the plant was able to significantly cut scrap at its extrusion line, reducing approximately $140,000 of wasted material in 1999 to between $12,000 and $15,000 in 2005. The plant ended 2005 with a gross scrap rate of about 1.1%. Its 2006 goal is 0.85% and as of April was on track to reach that mark, according to Funke.
"I think the big thing is we have frequent metrics that get in front of the people who run the equipment at least once a week, so they see these metrics as far as what are the major drivers, they understand what the cost of a quality defect is or what the inherent scrap is, and we use the Pareto 80/20 rule -- that we'll just focus on the big hitters until we can't theoretically drive anything out any further," Funke says.
