- THE MAGAZINE
- WEB EXCLUSIVES
Every time a new quality scheme is announced, I prepare myself for the initial, inevitable onslaught of consultants ready to impart their expertise, for a price, and take me from the “dark” into the “light” regarding my company’s quality practices. Many manufacturers relate the same experience and ask about such programs, “What’s in it for me?”
This question is being asked because of all the attention now being given to measurement uncertainty budgets. Uncertainty is not a new concept; it is reaching from the laboratory to the shop floor. And the answer to whether an uncertainty budget is worthwhile to pursue is simple: If you don’t know measurement uncertainty, all you have is a reading, not a measurement. Now that the technical stuff is out of the way, or nearly so, it might be interesting to look at measurement uncertainty from a business point of view. So, notwithstanding 12 Sigma or whatever level the quality professionals are touting this month, let’s focus on the business case for knowing measurement uncertainty.
An uncertainty budget, like a financial one, can be used to predict outcomes. And given recent headlines about corporate misdeeds, a measurement uncertainty budget may be more reliable than a financial budget when it comes to predicting outcomes. An uncertainty budget can be used for considering measuring problems and equipment selection. It may help uncover that what was thought would make a difference in a particular measurement situation, for example, new equipment, is insignificant to solving the real problem. This can save money. For example, my company is regularly asked to quote on the highest grade of gage blocks a customer insists are needed for his application. When we discover his temperature control varies by a couple of degrees and that the tool he’ll use these blocks to set has a resolution of 50 microinches, or 1 micron, we begin questioning his decision. Had he put an uncertainty budget together, he’d realize that the extra money for higher grade of gage blocks would be better spent on an air conditioner or a better tool, and purchase a lower grade of blocks.
Another way the budget can save money in tool purchases is by expanding measurement choices. For example, the most accurate way to measure something often requires expensive equipment, but not always. Knowing the uncertainty associated with different tools could indicate that a less costly tool is more than adequate for the application. Conversely, uncertainty budgets may show that only the best measurement tool will do the job, so spending money on anything else will ultimately be a waste, especially if rejects continue. And, the right tool will eventually have to be purchased, if the customer hasn’t cancelled the contract because of bad parts sent to him in the interim.
An uncertainty budget can help bring product engineers in touch with reality. Should your budget reveal that product engineers are asking for unrealistic part tolerances that are near impossible to measure within acceptable uncertainty values, it will become evident something has to give. A well-constructed budget may force engineering to prove they require the tolerances they call for on a part feature.
If a measurement dispute arises with a customer, comparing uncertainty budgets can help take personalities out of the conflict and reduce it to numbers. This allows a manufacturer to stand his ground without getting a customer angry. It also means there is no need to digest wrongfully rejected product. Conversely, it may show that something was overlooked in a budget.
There are standards that are followed to produce an uncertainty budget and tests that can be done to verify the adequacy of it. If a measurement uncertainty budget is done incorrectly, it costs in the outcome sooner or later.
An uncertainty budget should never be confused with governmental financial budgets where the rules are written to suit the situation and are rarely close enough to be used as a predictive tool. The only similarity between the two types of budgets is you pay when either is wrong.