A number of recent conversations reminded me how often a simple measure like OEE is misused and misunderstood. Since OEE can be very useful, a recap might be in order:
The purpose of OEE (Overall Equipment Effectiveness) is to provide a single measure that indicates how much good output we get from a particular piece of plant or equipment. In fact it’s a compound of three separate measures, all multiplied together to arrive at an overall percentage figure. These three measures are: Availability, Performance and Quality. Let’s look at each of these in turn, and then consider how to use the information.
Availability: of the total time that the plant or machinery could be running, what percentage of this time does it actually run? Time is eaten up by maintenance (planned and unplanned), break times for operators, set-up’s and changeovers and other reasons. Think carefully about how you will specify the total time: there are 168 hours in a week – will you use this as your baseline or the 7 1/2 hours per day for five days that is your normal working week? If you intend to run the machine for only part of the time or “as required” would it be sensible to use a “planned hours” figure as the baseline? Another couple of things to think about: most of us are familiar with muda, the Japanese word for waste, but not as many people think about muri or “overwork”. Overworking people or machines is counterproductive. Running plant and machinery 24/7 can disproprotionately reduces its working life and increase the likelihood of breakdown. Increasing availability means looking at improved scheduling (can be as simple as arranging operator breaks so that the machine isn’t switched off for lunch), reducing set-up and changeover times (using SMED), and introducing Total Productive Maintenance (TPM).
Performance: another measure that sounds simple but can often provoke argument. The idea is simply to compare actual speed or run rate with the “ideal” or rated speed. There can be a lot of “accepted wisdom” (usually B***S***) about “maximum” speeds – the simplest improvement is sometimes to simply crank it up! In complex situations where there are many interacting variables, consider using Design of Experiments to establish the optimum running conditions. Where possible, replace “adjusting” by “setting” and consider “centre lining” – establish mid-point or centre-line settings, to minimise the effect of “drift”.
Quality: here we’re looking at “Right First Time”, “First Time Through”, “First Time Pass”, Rolled Throughput Yield or similar – what perecentage of “perfection” are we acheving at the first attempt?
So now for some general points about using OEE:
Only measure OEE if you need to improve it – usually when you need more output. Remember: measure to improve – “you don’t fatten a pig just by weighing it!”.
The main purpose of an OEE measure is to improve the output of a specific piece of plant and equipment over time. It measures good output and identifies exactly where to improve. Concentrate on this rather than on pointless comparisons with other plant and equipment, other factories or other industries.
Generally, we always recommend that you use “tough” measures – compare where you are against the maximum possible or “ideal” situation, don’t simply work within your current constraints. It’s always easy to fudge the figures – don’t! Far better to improve from a true 30% to a true 35% than kid yourself that all’s well at an “untrue” 85%.
So – take a close look at those bottlenecks, measure the OEE and use your problem-solving skills and Lean tools to improve things!