Process Capability: The Heat Coil

Excerpted from Juran, Joseph M., "Architect of Quality - The Autobiography of Dr. Joseph M. Juran", Tata McGraw-Hill Publishing Company Limited, New Delhi, 2004, pp.97-101.

During 1926, one of my assignments concerned the "heat coil," a tiny circuit breaker that Hawthorne made by the millions. The sensing element was a wire-wound coil. A circuit overload would overheat that coil and cause the circuit to open. The inspectors regularly scrapped about 10 percent of the finished product because electrical resistance was out of specification limits. The local production supervisor and I became a team trying to reduce that waste.

The wire from which the coils were made was bought from outside suppliers and came in reels, each having enough wire to make several hundred coils. I measured many reels of the supply wire and found that although the electrical resistance was very uniform within any reel, the reel-to-reel variation was considerable. We got rid of that variable by measuring out from each reel a length of wire whose resistance was half way between the upper and lower specification limits, then cutting the entire reel into pieces of that same length. Now, despite the reel-to-reel variations, all lengths of wire were uniform in resistance. To my surprise, the defect rate dropped only slightly; nearly 10 percent of the product was still being scrapped because electrical resistance was out of specifications.

I next measured the resistance of fifty finished circuit breakers. Figure 9-1 shows the resulting "frequency distribution."


Figure 9-1 makes clear that the variation in the resistance of the final product is small in relation to the distance between the specification limits. But the average resistance of the finished product is not halfway between the upper and lower specification limits; it is distinctly lower. Why? A close look at the manufacturing process found the reason. The workers were soldering the wire to the units at a point several inches from the end of the wire and then breaking off the excess. The remedy was easy; we added several more inches to the lengths of the wires and presto, a defect-prone process became virtually defect-free.

I no longer have the original data from the heat coil project. In my second book (Juran, 1945)1, I gave an account of the project, recalling the data as well as I could.

I was intrigued by the ease with which we had added nearly 10 percent to the production of heat coils without adding any machinery, workers, or material. I knew from my experience as a troubleshooter that Hawthorne harbored myriad chronic defects. My imagination soared. Why not comb through the factory and get rid of every single one of them?

I discussed the idea with Vacin, my then department chief. He didn't share my enthusiasm, but neither did he reject the idea; he was simply indecisive. He finally suggested I talk about it with his superior S. M. (Smo) Osborne.

Osborne agreed that it would be wonderful to get rid of those wastes, but he pointed out a fatal obstacle. Making the product right in the first place was not the job of the Inspection Branch; it was the job of the Operating Branch. "They don't want us to tell them how to do their job, just as we don't want them to tell us how to do our job." So I found myself blocked by an organizational structure that made no provision for quality improvement.

The heat coil was the first instance in which I had quantified the inherent uniformity of a production process -- a property that today we call "process capability." As a byproduct of investigating the heat coil, I had stumbled onto inventing (or reinventing) one of the most useful analytical tools now available to practitioners in managing for quality.

Reference: 1. Juran, Joseph M., Management of Inspection and Quality Control (New York: Harper & Brothers, 1945).