Best Practices -- The Hunt For Red X

Maytag plant turns to Shainin techniques to reduce noise, vibration.

What to do with a washing machine that takes a walk when it's supposed to be working. Or one that won't shut up.

To ensure that customers won't have to suffer such annoyances, the management team at Maytag Corp.'s Herrin, Ill., plant turns to the legacy of Dorian Shainin, creator of the statistical engineering techniques that spawned the problem-solving tools known as the Shainin Strategies. Shainin, who died in 2000, founded Shainin LLC, a consulting practice that uses strategies such as Developed Lot Plot, Developed Precontrol, Variable Search and Component Search to solve problems. Shainin observed that for any one problem, the number of sources for that problem was infinite, but that one thing (dubbed "The Red X") inevitably accounted for nearly all of the undesired effect.

The Maytag Herrin plant, an IndustryWeek Best Plants finalist for 2003, turned to Shainin specialist Keki Bhote in the mid-'90s to get to the root of unacceptable noise and vibration in its washing machines, says Mike Tompkins, the plant's director of LeanSigma. Bhote and the Maytag team used Component Search to solve the problem.

Tompkins explains: "In the case of too much noise, we would take an exceptionally quiet machine and an exceptionally noisy machine. You rank the components you want to look at as potential causes of problems. The process is to disassemble and reassemble twice to see if there are any assembly issues with the components. Then you start doing swaps between the machines of individual components or subassemblies." By comparing the components one-by-one, it's theorized that the cause of the problem (The Red X) can be spotted.

"In one particular instance, we had a problem with a knocking noise in the transmission that turned out to be small pieces of aluminum chips from the machining operation [of the aluminum housing]. If that wasn't thoroughly cleaned, they could embed themselves in the gear teeth, and it made a loud knocking noise. Once we started the process, it took us less than a day [to find the problem]."

Time savings is a huge advantage of Component Search, Tompkins says. According to Jeffery Stevens, the plant's project quality engineer/LeanSigma master black belt, Component Search can save weeks of time and take far fewer resources than traditional Six Sigma processes. "One person could do a Component Search himself and locate the problem within a day or two." Additionally, training takes only a couple of hours.

Conditions need to be right, however, Stevens says. The defect must be defined (i.e., trying to eliminate an unacceptable noise verses designing quieter machines); the defect should be present in one "subject" and absent or acceptable in another; and the defect must be measurable.

"The accuracy of your measurement system should be about five times the width of your specification," Stevens says. Although it doesn't work in every problem-solving situation, Component Search has become a valuable tool in the LeanSigma tool box for the plant. In fact, the team is using it now to find out why some machines with unbalanced loads "walk" while others stay put. Stevens says the most time-consuming aspect of the hunt is identifying the test subjects. Otherwise, he expects a quick resolution to the problem of wandering washers.

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish