Not long ago, a manufacturer of needles for medical syringes was in the middle of another shift of quality assurance testing. The test involved three individuals side-by-side, with each grabbing the needles off the line and physically inserting them into a drill “chuck,” piercing a rubber slab to simulate the piercing of human skin. Transducers captured the force required for needle insertion compared to a benchmark measurement, assigning a “pass” or “fail” grade based on relative level of sharpness.
And then a strange thing happened. Seemingly out of the blue, one of the three needle testers began to get measurements that were way off the readings of the other two, causing all of the products passing through her station to receive a failing grade. Throughput took a huge hit. Supervisors were baffled. When you watched the three go about their work, everything looked the same. Their technique appeared identical, as did the amount of force being exerted to pierce the chuck. What was going on?