Overall equipment effectiveness (OEE) – What’s the point?

Anyone who is accountable for improving manufacturing performance should be skeptical about OEE. Without an inquisitive mind setting thoughtful goals, the outcome will likely not be success. In truth, OEE is a scoreboard and we are all playing the game. Every player and spectator needs to hold the organization and the officiating to account and make sure the game is a true exercise in improving performance. The reality is that countless conversations about how the game is played and how the matrix of statistics relates to the players will make a more successful enterprise.

It easy to get caught up in some academic implementation of OEE and lose the purpose of the endeavor. Let’s not do this. When asked to establish OEE in an organization, a large sum of money is often spent on expensive software and the expectation is a once-and-done implementation. Software will be important in achieving a practical implementation of OEE; however by itself, OEE is not software. OEE can be done with some worksheets and a clipboard. Everyone who is accountable, in regards to performance on the factory floor, needs to understand the numbers – that’s the point. When the numbers seem unfair or irrelevant to the model that represents the baseline performance, users should be complaining about the software and changes should be made.

I would like to challenge you to come up with a simple example that will allow you to integrate the concept of OEE. For example, I will track something simple.

Let’s assume I am a machine, outputting keystrokes. My output is determined by how many keystrokes I produce minute while I am working. I am scheduled to work 8:00 to 5:00, five days a week. For simplicity sake, let’s say I will type all day long and be fully loaded (9 hrs.). I take an hour lunch and random amount of distractions take me away from the keyboard and effect the availability (5 hrs. at keyboard) of keystroking. I have a skill that has some level regarding hitting keys, a possible performance (WPM 22) expectation of so many words per minute I may produce. I may not produce a quality (3 out of 22) results with all this keystroking so good vs. bad content needs to be determined.

This below example shows the simple OEE web application OEE calculation based on the above data

I was not typing for the whole shift, so I had 5/8 (0.625) of the day available in at 5 out of 8 hours. Which is my loading for the day minus my hour lunch break.

I could have produced 22 (WPM) * 5 hrs. 5700 total words. I fell short at 5000. Making my performance less than 100%.

I produced bad quality words upon review. This is separate from performance and reflects only ratio of good verses total pieces.

What’s the point? This is really just more data. To produce meaningful information we need to know what the baseline is. Basically, what the history is. Just like a baseball statistics like, Times at Bat, Hits, and Bases Run it gives an expectation to all the stakeholders we can mark improvement against. It give us data goes deeper than what is currently happening, incorporating the history of data by tracking stats. Creatively applying knowledge to the way the game is played based on current conditions is the point. Without the critical input and the ability to change the current trajectory or events this exercise would be somewhat meaningless. In the end the players need to standup and want to play ball.

There’s plenty to argue and disagree about in the above example. But, by having these arguments we are setting our focus on overall performance.

1. Who to say that my maximum words should be set at 22 WPM. That rate is not all that impressive. Also, some words are harder than others? Setting performance expectations too low may inflate numbers. We should never hit 100%.

2. One could argue that someone should have kept typing on my keyboard over lunch if I wasn’t typing. Maybe I did type over lunch but counted me break anyway. Under stating availability by making claims of scheduled downtime can inflate numbers as well.

3. How useful are these words are they quality words? This assumption can be hard prove. Overlooking bad quality could inflate OEE numbers. If I sent these words to a publisher and they were all rejected my true cost could be much more than doubled if I need to replace the words.