Add Value with Measurement
August 27, 2007
One of the jobs of metrology system manufacturers is to make better, more sophisticated products. Ironically, however, they are frequently met with the argument that because inspection is essentially a nonvalue-added process, their efforts in product development run counter to their customers’ needs. Their customers need to reduce costs, and inspection is a cost. Ergo, the price of high-end form and surface metrology systems-and the development of ever more sophisticated ones-is not justified. Instead, they are told, they should concentrate on simpler, more accurate, reliable and less expensive gages for inspection.
It is an interesting argument, and one not without merit. Inspection is essentially a nonvalue-added process-a manufacturing cost-and companies should generally strive to do less of it. Indeed, a central tenet of Deming’s philosophy of statistical process control was to reduce the need for part inspection by controlling the processes that lead up to it. But while I am in favor of simpler, more accurate, reliable and even less expensive gages, that does not argue for a reduction in the ability to measure. It argues for more capability.
easurement, as opposed to inspection, is a value-added process, and from a metrological point of view, the ability to make the right kinds of measurements at the right time can reduce the need for inspection. It also can improve products and the processes that make them, and ultimately, improve profitability by driving the overall efficiency of a business.
Inspection vs. MeasurementIt is generally agreed that some form of measurement or inspection is critical to producing a quality product. But not all measurement is inspection. According to the dictionary, the verb form of inspect is to look something over carefully and critically, especially for flaws. But to measure is to ascertain extent, dimension, quantity and capacity, especially by comparison with a standard.
So measurement is a much broader concept. Inspection means someone created something and now wants to see whether it is good or bad. Inspection always answers a yes or no question: is it good or bad, pass or fail, go or no-go? So if operators spend all that time and money to make something, get it to a finished state, and then find out-oops-here is another bad one, then they have wasted all that money.
Process control helps address this problem and is generally acknowledged to be highly value-added. But if one takes that concept a step further and uses appropriate measurement to design a robust process that requires a minimum of control, they reduce costs for both statistical process control (SPC) and inspection. Go another step and use appropriate measurement in the product design process, and one reduces inspection costs even further by choosing the most reliable design, and setting tolerances as broad as possible, while ensuring functionality of the product.
The value of measurement lies in the types of decisions it allows operators to make. Measurement provides data-information as opposed to yes/no-and information informs decisions. An operator can decide to manufacture one design over another because one is more robust. Or, they can decide to manufacture using one process instead of another because it requires less investment in machinery, or involves one less operation or produces a tighter tolerance.
Paradoxically, in many companies, the oldest equipment is often in the R&D center because no chips are cut there and nothing is really made there. However, by making appropriate investments in equipment and by pushing metrology further upstream to provide better product and process design data, downstream costs can be dramatically reduced. And not just inspection costs. When thinking about improving measurement capability in a company, think about the entire enterprise.
Business ManagementSay someone is running a business. On one side of the ledger, they have income from sales. On the other, they have expenses for people and materials, and operating expenses for things such as telephones, heat and electricity and, of course, amortized capital costs for equipment. When comparing the two, what is left is profit. A business manager wants to maximize profit.
By investing in the ability to get the correct measurements, the impact of all these expense categories can be minimized. If inspection is reduced, not as many inspectors are needed, thereby reducing personnel costs. Without the same high scrap rates, material costs are impacted. And if not as much money is spent on inspection equipment, operating expenses are reduced.
So if someone wants to make a case to the manufacturing manager who got his MBA somewhere where he learned that inspection is nonvalue-added and does not really understand why he should do something that looks a lot like inspection to him-because inspection is a bad thing-talk in terms that he will understand. He needs to understand how making the correct type of measurements can favorably impact the business.
No one would think of running a business without other types of measures. How long would the CFO last, if when asked, “Are we making a profit?” he responded, “I don’t know. We’ll have to wait till the end of the month to see if there’s any money left in the account.” Companies measure all types of things, such as inventory and receivables, vacation days and on-time deliveries. So they should not run a business without data on quality, products and processes.
Practical ConsiderationsThese are not new ideas, of course. More than 400 years ago, Galileo said, “Measure what is measurable, and make measurable what is not so.” And it was Lord Kelvin in the nineteenth century who postulated, “When you can measure what you are speaking about and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge of it is of a meager and unsatisfactory kind....” He also said, “If you cannot measure it, you cannot improve it.”
This last statement is one that many people in quality roles today will find hits close to home. Improving quality is fundamentally about understanding a process, gathering data to confirm theories and driving improvement based on this data. Having the best possible data with the lowest uncertainty of measurement is important to making the best decisions.
What is meant by a value-added measurement? As stated previously, the value of a measurement lies in the value of the decisions that can be made with it. It is not quite so useful if one is just getting a number to determine pass/fail, or compare it with the tolerance of a similar part.
An example of a value-added measurement is measuring two different produced or designed components and ascertaining a differential value for each. If one component does not work as desired and the other does, the operator has just established a functional tolerance. Using measurement data to set tolerances to the maximum allowable while still ensuring a fully functional component is an example of a valuable measurement. Having a better tolerance value ensures that the thousands of components that are made later have the benefit of the largest tolerance possible.
Another example of a value-added measurement is using measurement data to evaluate two different manufacturing processes for a component. If it can be confirmed through the use of the data that a less expensive process can be used for manufacturing the component, tens or even hundreds of thousands of dollars can be saved.
Getting real numbers is important. Inspection is minimized when designs and tolerances are optimized and when the correct process has been defined. The earlier in the process good measurements can be done, the better off a company will be. Driving the measurement process upstream means less work to do later. Instead of measuring 3,000 components during the production process-or controlling an unoptimized process to do so-only about 20 prototypes will need to be measured.
However, in the constant pressure to compete and the push to get products to market, other things too often take priority: there is no time to establish functional tolerances on prototypes, and there is a belief that the process can always be optimized later. But as numerous studies have shown, most of the cost of a product is determined very early on in the design process. After the design is chosen, the manufacturing process does have some impact on the cost, but the options are limited by the design. Finally, after the process is chosen, later attempts to optimize it can only make minimal improvements to the cost.
Quality professionals can all do their part to make sure that they help their companies make value-added measurements. They need to make sure that the tools are available to develop the best products and processes, and that they take the time to use them. Companies that do this will consistently minimize cost and improve their competitiveness. Measuring the right things at the right time will add value to products and processes, and ultimately, help the business grow. Q
- Measurement is a value-added process, and making the right kinds of measurements at the right time can reduce the need for inspection.
- Measurement can improve products and the processes, and ultimately, improve profitability by driving the overall efficiency of a business.
- The value of a measurement lies in the value of the decisions that can be made with it.