The Overall Equipment Effectiveness (OEE)
How to understand, measure and implement it
The OEE (Overall Equipment Effectiveness) is a measure of the goodness of use of a machine (or a process) to its full potential. It’s a “best practice” to monitor and improve the efficiency of a manufacturing process (eg: machines, work cells, assembly lines). The OEE encompasses all the most common sources of productivity losses in three factors and is used as one of the most critical metrics in TPM (Total Productive Maintenance) and Lean Manufacturing because it provides a robust and shared method for calculating efficiency.
The OEE analysis starts with calculating the total opening time, that is the amount of time in which the factory is open and in which the machines can work. From this amount of time, the Planned Downtime is subtracted, which includes all those events in which production is not scheduled (lunch breaks, planned maintenance, periods in which there is nothing to produce) obtaining the available Operating Time.
For the calculation of the OEE, we start from this calculated time and divide all the efficiency losses into three categories: Availability, Performance, and Quality.
Availability
Availability takes into account all those events that stop the planned production for a period (usually exceeding ten minutes) and is calculated as:
Availability = (Total Operating Time) / (Available Operating Time)
Examples of events that end up in these unplanned downtimes are: machine failures, material shortages, and model change setup times. Although setup times are necessary, they are counted as a waste of time because they can be reduced, for example, by applying the SMED technique.
Performance
All speed losses are counted in the Performance factor which is given by:
Performance = (Net Operating Time) / (Total Operating Time)
All those events that cause the machine to fail at its nominal speed are classified as speed losses, for example, wear of the machine, sub-standard raw materials, and inefficiency of operators. The Net Operating Time is usually calculated by taking the nominal or theoretical cycle time of the machine and multiplying by the pieces produced in the time considered:
Performance = (Ideal cycle time x pieces produced) / (Total operating time)
Quality
The Quality factor takes into account all the time lost in the production of defective parts because they do not reach the desired standards including parts that have to undergo rework. It is calculated as:
Quality = (Value Added Operating Time) / (Net Operating Time)
and it is practically calculated considering the number of good parts to the totals produced in the period considered:
Quality = (Good Pieces) / (Total Pieces)
The OEE and the “6 Great Wastes”
The OEE takes into account the three factors presented above, the calculation formula was proposed by the Japanese engineer Seiichi Nakajima in 1960:
OEE = Availability x Performance x Quality
The primary objective in OEE calculation is to reduce and eliminate the “6 major wastes”, that are the most common causes of inefficiency in manufacturing processes.
Once the losses and their sources have been categorized, we focus on monitoring and correcting them. With the sensors available today it’s possible to make a quick and real-time classification of the main causes for each category.
Here is a brief discussion of the possible improvement methods to solve the “6 Great wastes”.
Faults
Eliminating downtime due to breakdowns is essential to increase OEE, however, it is important to know when they occur and the cause that originated them. Once all this information has been tabulated, a “Root Cause Analysis” is usually applied to identify the root cause of the problem.
Setup and adjustments
The times spent in Setup and Adjustments are usually measured considering the time that elapses between the last good piece and the first good piece between a mould/model change. This time often includes adjustments and ramp-up times. Measuring and verifying setup times over time is essential to reduce this waste and must be combined with a setup time reduction program such as SMED.
Short Stops and Reduced Speed
These types of wastes are the most difficult to track and solve due to the difficulty in their registration and classification. Short stops are usually considered to be stops of less than 5 minutes that do not require the intervention of the maintenance department. This threshold is chosen because the reason for the stoppage both in the case of manual data collection and through electronic sensors must be filled in / entered by the operator in charge of the process; understandably, the stress that would be created in having to continually enter reasons for stoppage for an unstable process would affect the work of the operator. For this type of loss, “Cycle Time Analysis” should be used, in which thresholds are established to distinguish between Short Stops and Reduced Speed by comparing the cycle times collected with the nominal one. Even if the effects are very similar, short stops and speeds below nominal are divided into two different categories because the root causes are generally very different.
Rejects in restarts and rejects after production has started
The rejects are divided into two categories according to whether they occur with the machine starting after a shutdown/setup or with the machine operating at constant speed; this distinction is made because the causes are generally different. The parts that require rework must be considered waste, in fact, the OEE measures quality from the perspective of the “First Pass Yield”, ie the number of good units that come out of a process compared to the number of units entered. Tracing the moments in which wastes are produced within a shift or a working day helps to identify the causes and often leads to the discovery of recurring patterns that originate these wastes. Often Six Sigma programs where one of the target metrics is achieving fewer than 3.4 defects per million opportunities are used to achieve near-perfect quality.
Disadvantages of using OEE
In a company that pursues Lean Manufacturing, it’s essential to develop key measures to support the behaviours that create value and to ensure that everyone in the organization can measure and be evaluated for the efforts that lead to a reduction of waste in processes. If an organization remains tied to the old productivity measures typical of mass production, it will be difficult for workers to change their values and behaviours. In this sense, the OEE is a perfect metric in the evaluation of productivity improvement programs, however, if misinterpreted or applied in certain situations it can lead to making wrong decisions.
Use of the OEE in a pacemaker process
When the OEE is used on a pacemaker process where the cycle time is defined by the customer’s request and not the rated speed of the machine, the Performance component can be very low. In fact, in a pull production system, the pacemaker process gives the input for upstream production and if it’s made to work at a speed higher than the takt time, overproduction occurs, the first of the wastes identified by Ohno, while at a lower speed you will not be able to deliver to customers. In cases like this, the customer’s takt time should be taken as the cycle time for the calculation of the OEE, however, since the takt time can vary monthly, it is difficult to choose a stable cycle time that allows the historical comparison of the OEE to verify the trend.
The aggregate data
Another problem in the use of the OEE can arise when you look only at the aggregate number without considering the three factors of which it is composed. If we take for example a process that presents the following values in the first turn:
Availability 75%
Performance 95%
Quality 95%
OEE = 75% x 95% x 95% = 67.7%
While in the second round:
Availability 90%
Performance 98%
Quality 85%
OEE = 90% x 98% x 85% = 75%
One might think that there has been an improvement in the performance of the machine, which presented greater availability and maintained a higher speed but at the expense of a deterioration in quality which in economic terms is usually more expensive for a company. Looking at the total data can therefore lead to a diametrically opposite behaviour compared to Lean production, whose objective is to use the machines when requested by the customer and at a speed such as to produce parts according to quality standards decided by the customer.
What the OEE does not show
The OEE is a good way to represent the efficiency of a single machine, but it can lead to incorrect results when evaluating the efficiency of an entire production site. The plant OEE can be calculated by making a simple arithmetic average of the OEE of the individual machines or by making a weighted average based on the planned working time on each machine. The fact of removing the unplanned time from the total time available can lead to higher OEE in times of low order intake from customers and therefore to the planned downtime of some machinery. In this case, it is also useful to accompany the OEE with a measure of the use of the machinery, for example:
The use of assets, the ratio between the current operating time and the total time available (calendar days, 24 hours a day)
Utilization of capacity, the relationship between production time loaded and the time available.
Enrico Mantovani
- If you enjoyed this, please follow me on Medium
- Interested in collaborating? Let’s connect on LinkedIn
- Please don’t forget to clap for the article :) ← Thank you!
