The traditional formula for overall equipment effectiveness (OEE) — including speed, quality and uptime — treated all three factors as equals and OEE as a measure of equipment effectiveness over a 24-hour period, assuming 24 hours was the required performance window.

Most companies have refined the calculation to fit the actual required performance window. Robert Williamson’s example was the window when a racecar was required to operate at peak performance. In your plant, it may be continuous operation for only one shift.

The metric began to be useful as a plant-floor performance measurement and, from analyzing the three factors, an analytical tool to determine improvement opportunities. All three factors are really at the mercy of the operator, who is the most important person in the plant when equipment is needed.

I submit that all three factors (speed, quality and uptime) are not necessarily one-to-one linear correlations. Does a percentage increase in speed maintain the same quality or a proportional decrease in quality? How about the relationship to downtime?

More importantly, are the savings associated with a 1-percent increase in speed offset with the reprocessing costs of, say, a 1-percent decrease in quality? I found this factored into the formula by a plant that penalized the formula seven-fold for each percentage change in quality. In my question about the 1 percent, they had figured the quality-cost impact to be equal to seven times the savings of a 1-percent increase in speed. Therefore, the OEE formula was impacted by a quality change of 7 percent for each actual 1 percent in quality. Quality was indeed “job one.”

Have you looked at the optimum OEE factors for your equipment in terms of costs associated with increases and decreases in the factors? In other words, do your total processing costs necessarily go down with an increase in OEE?

I have come to the conclusion that all involved must understand the impact of any change in an OEE factor in dollar terms. OEE for the sake of OEE can bankrupt you. Decide on the range of performance of each OEE factor that can be tolerated, and educate all concerned on how that was determined and the costs to the process from changes outside those ranges: rework, scrap, upstream and downstream performance, maintenance, consumables, work hours, overtime, etc.

The biggest loss of continued games with OEE is the trust, confidence and morale of the employees. They are not blind to the effects of “increase the speed by 10 percent and we’ll get out quota tonight.”