The Adventure of the Shifty Weights

A ChemE instructor uses this case to teach students that the answer is to an engineering problem often is literally right in front of you

By Marv Havens, Contributing Writer

Over forty years ago, I took the Introduction to Chemical Engineering class with Dr Harvey Grice. He had retired from decades with industry and made it his mission to teach us young students the practical side of engineering. His favorite technique was to tell war stories and bring them to a point. One lecture still stands out after all these years:

Dr. Grice brought in his work notebook from a stint at Birdseye Foods. They had been having a quality control problem on one shift, basically frozen pea package weights would run heavy for a while, then drop and run light, a step-function without any apparent time period. The weight variations were roughly the same magnitude, just heavy or light of the tare weight.

He added that the operator’s job was to pick a package every few minutes and check its weight on a tared mechanical balance. The operator would then adjust the filling operation for more or less weight as needed. Further down the line, the QC department also did a weight check and plotted the data. He gave us a plot of the QC department’s data and asked us to suggest causes and solutions.

After some discussion, he offered his observations and answer:

Remember that this was decades ago with mechanical balances, meaning a needle against a scale. The balance was tared such that at the correct net package weight, the pointer would be straight up. It seems that on the shift in question, the operator was a bit, uhm, large. She sat on a small stool, “parking” herself to one side. She would then pick up and weigh packages, and adjust the filling operation for what she saw on the needle reading.

The process would run along for a while on this adjustment. Then she would shift position to the other side, take more packages and adjust the filling operation accordingly.

Parallax was the problem, depending upon which side the operator sat on. There was an offset (either positive or negative) that correlated with the angle she viewed the needle from. So the filling operation would run heavy for a while, then light, at always about the same offset and with no specific periodicity.

The solution: A bigger cushion on the operator’s stool. And possibly a change in diet that involves more vegetables.

Dr. Grice made the point of his war story a teaching opportunity: For most engineering problems, the answer is literally in front of you. As engineers, our job is to recognize it. Once seen, the resolution is often straightforward. And his lesson is still pertinent today — you cannot solve engineering problems from the office. Get out there and see what is real.

Thanks Harvey. I try to pass on what you taught me.

Contributing Writer Marv Havens is a BC (Before Calculator) Engineer. He is currently a chemical engineer at Sealed Air’s Cryovac Division in South Carolina. Previously, he worked for Shell Oil and Union Carbide. He has thirty patents and teaches in-house classes in Polymer Science. He still contributes to a scholarship fund in Harvey Grice’s name.