SCIENTIFIC QUALITY MANAGEMENT

"The activities grouped under the umbrella label 'quality management' often constitute the greatest opportunity for reliability and efficiency increases"

We define quality as meeting the specifications provided to the buyer at the time of purchase. The specifications must accurately describe the function of the product, the environmental conditions in which it will work, and the length of time the owner can expect the product to work without fail.

Quality alone may not be enough
Quality alone does not always drive sales. An unsold puce–colored new car that stays on the dealer’s lot is of no lesser quality than an otherwise identical blue one that sells instantly. And quality is often confused with value in which two products perform identically but one is priced lower than the other. Which may explain some of the confusion in the position that “quality is what the customer says it is.” (All of this gets rather involved; for a fuller explanation, we suggest the landmark 1996 book Optimizing Quality in Electronics Assembly: A Heretical Approach by our Jim Smith and Frank Whitehall, formerly director of corporate quality at Edinburgh’s Ferranti Defense Systems.)

Given the choice between two similar products of identical quality but different prices, a rational customer will pick the lower priced product. Therefore, the way to maximize the company’s profits — and profit maximization is the ultimate responsibility of every department — is providing highest quality at lowest cost. If the quality department fails to reduce costs, the company suffers.

Quality and documentation
Documenting a process (in accordance with ISO/TS/TL) is generally the necessary first step towards eliminating defects in output. But it is only the first step. It can be tempting to confuse paperwork for the goal rather than a tool. If the documentation becomes too dense, usability suffers. In extreme cases (especially in the automotive, telecom and aerospace/defense electronics), we have encountered process engineers unwilling to undertake process improvements because the existing process documentation has been blessed by a customer representative.

Process documentation must be concise and flexible. Changes to the documentation must be expected and encouraged. We approach the challenge of documentation in the same manner we tamed reliability criteria (itself a critical element of quality management).

About data
Every competent quality professional knows that quality achievements (and, for that matter, process efficiencies) begin with data. However, knowing that data are important is not the same as knowing how to gather and use data effectively. Understanding what to measure, where to measure, how to measure and what to do with the measurements requires a great deal of process knowledge to start with.

From our earliest days, we have specialized in setting up data systems that are easy and effective. At the most basic level, this means:

    • Recognizing that every defect reflects a defect in the process; defects are symptoms of a sick process.
    • Ensuring that defects are accurately categorized and uniformly recorded; many well intentioned quality systems have foundered simply because different inspectors using different terminology for the same issue.
    • Disciplined recording; give an inspector a soldering iron to touchup output and incomplete recording is inevitable.
    • Identifying the location of defects; seeing a defect occur time and again in exactly the same place has different meaning than defects occurring randomly.
    • Timely feedback to production; reports generated even a few hours after production deny process managers the opportunity to find what changed in the process.

Quality systems that require constant universal inspection/touchup/rework are failed strategies. Quality data exist to drive process improvements and those improvements must drive steady reductions in the amount of monitoring needed. The number of defects must be so small that the odds of discovering them even through 100% inspection are insignificant. At that point, first piece checks and random reviews are employed to guard against the production processes imploding through some freakish event.

Quality, process engineering and production
The most successful quality systems tightly integrate quality, production and process management. Quality sets the requirements and compiles data collected in the course of production. Process engineering then employs the data to strengthen the process.

The least successful quality systems operate at arm’s length from the actual production. In those situations, quality essentially grades the production and process engineering. Inevitably, such policing creates resentments and covert rebellions.

Root Cause, corrective actions and customer education
In market segments with a few powerful customers, quality departments typically manage resolution of customer complaints. The complaints can range from the profoundly important (failures) to the unfounded (frequently involving a customer inspector unhappy about a solder connection that actually conforms to the specifications and real performance requirements).

Failures require effective root cause analysis and corrective actions. We have a long and successful track record helping clients with both steps. (More inspection is not a corrective action; it signifies inability to solve the real problem.)

Acquiescing to unfounded demands, rarely ends well for the supplier or the customer. (Customers don’t say “that’s a good supplier because it replaced the parts we didn’t like;” rather, they say “that stupid supplier would have caused us great harm if we hadn’t been so alert and caught the bad units.”) Often, we educate the customers as well as our own clients.

Expert input
Sooner or later, all electronics manufacturers suffer product failures caused by supplier error. Generally, the suppliers will deny responsibility. We help our clients prepare and press the case for restitution of damages. But we also help them manage suppliers to minimize the risk of errors.

Quality is a science
We began this by saying quality is an odd business. It shouldn’t be. Like every aspect of electronics manufacturing, quality systems must be based on science. And that’s where we can help. For more information about how Electronics Manufacturing Sciences can help your quality systems, please send this simple form or call (01)727–866–6502 ext. 21.

Copyright