OEE: The Gift That Keeps on Giving

Combining key data can paint a better picture of what is happening in a shop and help it make decisions that will yield the greatest long-term benefits.

Overall equipment effectiveness, or OEE, is a metric that accurately conveys how well machines of all types are performing. More and more companies are finding OEE to be a key component in their continuous improvement initiatives.

As with any metric, OEE involves collecting data—specifically, data pertaining to equipment availability, performance efficiency and rate of quality. The meaning of these three OEE categories is described here: 

• Equipment availability is the amount of time a machine is available to produce parts when needed. This would be the time remaining after machine breakdown and setup times are removed. For example, if during an eight-hour shift a machine experienced breakdowns for one hour, and another two hours were spent setting up jobs, machine availability would be five hours. Expressed as a percent of available time, equipment availability would be 63 percent (five hours divided by eight hours).

• Performance efficiency is the percentage of time that a machine is running at planned speed. Factors that affect planned speed are speed reductions (that are usually related to the part being run) and minor machine stoppages (such as tool changes, material loading or an operator being called away for a short interval). In the previous example, if during the five hours that the machine is available to produce parts, it produces the required number of parts within four hours, performance efficiency would be 80 percent (four hours divided by five hours).

• Rate of quality is the percentage of time the machine is producing good parts. The time spent producing defective parts is deducted from the machine’s run time. Following our example, if 10 percent of the parts the machine produced during the four hours of productive run time were defective, the rate of quality would be 90 percent.

Using the numbers in this example, multiplying the percentages of equipment availability, performance efficiency and rate of quality would yield an OEE of 45 percent (63 percent multiplied by 80 percent multiplied by 90 percent). This means that the machine is running parts of acceptable quality at planned speed only 45 percent of the time, or approximately 3 1/2 hours in an eight-hour shift.

Interestingly enough, many companies already collect this data, but do not combine it to calculate OEE. Equipment availability, performance efficiency and rate of quality may even be reviewed and evaluated by different departments independently. Maintenance may capture machine downtime and track this as part of a daily, weekly or monthly metric. Manufacturing may capture performance efficiency in terms of whether machines produced the number of parts they were supposed to in a given period. Finally, quality control may collect data on defective parts produced and track it as part of a plant-wide yield rate. Unfortunately, all of this information treated individually may not present a clear picture of how a machine is really performing.

OEE ties all of this information together and can help a company make better decisions. One company recently set out to use OEE as the preferred metric for its key machines. Like many companies, it recorded lots of data, but in a rather disjointed manner. The company’s main focus had always been machine efficiency, and machines were ranked by the number of parts they produced per hour, when they were running. When the management team looked at machine performance from an OEE perspective, there were some surprises. For example, one machine, let’s call it the “fast” machine had a strong record of producing a high number of parts per hour and was considered 95-percent efficient. Another machine that did the same type of work, let’s call it the “slow” machine, could not produce parts at the same rate and was rated as 80-percent efficient. On the other hand, the fast machine had long setups and experienced frequent breakdowns, resulting in equipment availability of only four hours per shift, or 50 percent. The slow machine could be changed-over much faster and rarely experienced any breakdowns, so it had an equipment availability rating of 85 percent. Finally, good parts were produced on the fast machine 93 percent of the time, while the slow machine produced good parts 99 percent of the time. In the end, the fast machine’s OEE was calculated to be 44 percent (95 percent multiplied by 50 percent multiplied by 93 percent), while the slow machine’s OEE was a much higher 67 percent (80 percent multiplied by 85 percent multiplied by 99 percent). To put this in perspective, over an eight-hour shift, the slow machine was producing good quality parts at planned speed almost two hours longer than the fast machine.

Focusing on OEE, rather than individual categories of machine performance, can provide a better picture of what is happening with your machines and allow you to plan the changes that would yield the greatest long-term benefit.