Quality control tips: How to avoid making errors in the first place
While error-proofing isn’t free, it can be considerably less costly than shipping the wrong parts, causing a shutdown, or triggering fines or remediation efforts that are built into some supply contracts, says David Marten, Faurecia manager of manufacturing metal engineering.
While error-proofing isn’t free, it can be considerably less costly than shipping the wrong parts, causing a shutdown, or triggering fines or remediation efforts that are built into some supply contracts, says David Marten, Faurecia manager of manufacturing metal engineering. “If we lose a half day’s production due to either a machine being down or having produced unacceptable product, we’re in big trouble. We have to do error-proofing, or we’re not going to make it as an industry,” Marten advises.
Marten works with others on the subcommittee, Effective Error-Proofing Workgroup, under the Quality Steering Committee of Automotive Industry Action Group Forum (AIAG) . Faurecia makes automotive seating in Troy, MI, and AIAG is making a guide to error-proofing. In 2006 the committee began work; a draft of the guide is ready this month (December 2007). Final approval is expected in March 2008 leading to a published guide by July.
Faurecia's David Marten provided this error-proofing example, showing a rack where parts are picked. A green light will flash showing the correct location to pick the correct part. It verifies the correct part was picked, using safety type light curtains to verify the correct bin. A scanner (far left) is used to read the bar code to further ensure the correct part is used. If someone picks from the wrong location, a red light flashes, and the process is stopped until it is corrected.
And the answer isn’t to automate everything, because for some jobs with high variability and few final products, people can be more cost effective than alternatives.
“People are very good and very flexible in a lot of processes where controls that are too rigid or robotics just are not cost effective,” Marten says, “even though robotics have come down in cost quite a bit. We use a lot of robots in welding, especially. They’re better for a lot of things—for safety, fewer errors, and speed. Automation helps with error-proofing, but it doesn’t mean you eliminate operators. We just control the environment so humans have to do things right or the process shuts down.”
The future of process error proofing directs efforts toward design. The same error-proofing methods can be applied to new technologies or spin-offs of existing technologies, with subtle differences in application and sequences. Error-proofing should eliminate errors or catch them as soon as they occur.
Safety or warranty issues must get higher priority for resolution, Marten says. They must be addressed immediately with existing products or planned for elimination in new or modified products. For error-proofing within existing processes (reactive or proactive), determine the kind of data needed to begin the error-proofing efforts. Key tools are a design assessment, which include the following elements: ergonomics and safety, assembly related, part shape / color, part function, and part complexity.
Roughly 10% of project cost is in controls; and 50% of error-proofing methods or devices is in controls, Marten says. More parts mean less reliability. The idea is to improve design, so the product can only go together only one way, the correct way.
Does better quality control always increase costs? Not necessarily. Perhaps a better question is how much does it cost when there’s a problem? Keep corrections as simple as possible. Seek the most effective, least cost way of eliminating the possibility of error.
Among means of error proofing, from most desirable to least desirable are:
- Eliminate error prone element (product or process design);
- Prevent error;
- Detect error and prevent defect;
- Detect defect at source; and
- Detect defect prior to shipping.
The reason the last is the least desirable is because, like a corollary to Murphy’s Law, if defective parts are made, eventually they will be shipped.
Labels can be a good place to start. With today’s just-in-time manufacturing, there are just a few hours of extra stock, so there’s no leeway to ship or misidentify any stock, Marten says. Labeling caused many problems: Wrong container; new, old, or wrong parts in container; wrong count; wrong part label, label in the wrong location, and illegible label are among challenges. A label can cost 2 cents or less, but a labeling problem can cost is $5,000-$10,000 per incident, including containment and sorting, third-party containment, problem-solving meetings (usually with a lot of yelling), customer help, and related actions, Marten says. “Then you get someone in your plant to help, to watch everything you had planned on doing anyway,” he adds, with a slight smile.
If labels are printed ahead of time, the operator must identify the correct label, manually apply it, identify parts, and place the correct number of parts in correctly labeled containers. Most are date coded or shift coded, and sit around after the shift, leaving open the possibility of an error.
Ways to error-proof labeling include: Improve lighting, bar code parts, verify placement, handle one part at a time, print labels at points of use, print one label at a time and apply immediately, print container label only after the container is full of the correct quantity, physically separate like parts, any work in process (WIP) must have every part identified, with added verification for anything unusual or of special circumstances. Add numbers or geometric shapes to labels for easy identification, mold or etch bar codes. Keep colors and shapes for labeling consistent. (For instance, always use triangles to indicate the color “charcoal black.”)
Other fixes to help error-proofing labels include: Eliminate labels with etching, minimizing off line operations, prohibit dock audits, install sensors for label presence and box movement, use a printer with minimal operator influence, use color sensors and unique packaging, and develop controls to correctly label things 100% of the time.
Sometimes it’s just silly stuff that we do to ourselves, Marten says, recalling the time a quality inspector removed a part from a box for a test, and the box was shipped before the part returned. Now the inspector must pull parts before they get packaged.
Some aspects of error proofing can be counterintuitive. Being ready for the challenges can move the process along.
Questions that can come up when thinking about or discussing error-proofing changes:
- What don’t I understand about the products or processes?
- Why do we do it that way?
- Isn’t it too difficult or too expensive?
- Don’t I pay workers to do it right every time?
- Nothing is perfect, right?
Marten made the comments in a presentation at National Manufacturing Week in Rosemont, IL, near Chicago, in September.
— Mark T. Hoske , editor in chief
Control Engineering System Integration eNewsletter
( Register here and scroll down to select your choice of eNewsletters free .)
|Search the online Automation Integrator Guide|
Case Study Database
Get more exposure for your case study by uploading it to the Control Engineering case study database, where end-users can identify relevant solutions and explore what the experts are doing to effectively implement a variety of technology and productivity related projects.
These case studies provide examples of how knowledgeable solution providers have used technology, processes and people to create effective and successful implementations in real-world situations. Case studies can be completed by filling out a simple online form where you can outline the project title, abstract, and full story in 1500 words or less; upload photos, videos and a logo.
Click here to visit the Case Study Database and upload your case study.