The numbers game

Everyone in manufacturing knows about overall equipment effectiveness or OEE – or do they? Which of these is the right definition: is it an internal measurement of production efficiency? Or a way of assessing the efficiency of TPM (total productive maintenance)? Or a catalyst for improved ways of working? Or simply the one measure that highlights the opportunities your business has ignored? And should it be applied to a machine, to a line or to the entire plant? I've heard them all in the past month from people in very responsible managerial and technical positions. And let's just throw this into the pot – at a US manufacturing conference, 60% of the delegates didn't know how to calculate OEE. So, for a universal term, it's got a pretty elastic identity. It doesn't mean any of these definitions are necessarily wrong. Just that people see what they want to see. So let's take some expert advice. Dr Roger Benson has over 38 years' experience in global process industries. He is a fellow of the Royal Academy of Engineering, author of Benchmarking Process Manufacturing, and a judge of the UK Best Factory Awards (BFA) programme. What to him is so important about OEE? "OEE is the best measure of the operational efficiency of a plant. It applies to all industries without exception. Most pretend they are different but they are not." He points out that even though it sounds simple – availability x performance rate x quality – it is a very demanding measure. "Each of the three factors is measured at less than 100% so, measured together, you can get a pretty small number even though world-class figures are very high. It tells manufacturers what the size of the hidden opportunity is in their operation and it is also the number that gives you the biggest financial incentive." More of that hidden opportunity anon. It's worth taking another look at Benson's view of those three factors first; he has some astringent opinions on how they are calculated and what he calls the 'fiddle factors' seen in plants of every type. Firstly, availability: the hours a year the plant was available to actually produce, divided by the maximum possible (8760 hours). "The fiddle factors are people saying we don't run weekends or shifts or leaving out annual holidays," he explains. "These are just a distraction. Okay – it's nice that in Europe we have our five-week holidays but unfortunately the Chinese don't run the same way. I used to say to my colleagues at ICI, 'Let me run the plant at the weekend and I'll make own-brand paint.' Their reply was 'You can't do that!' If they don't want to run weekends or holidays, that's a commercial decision. But they are losing an opportunity in the plant and OEE quantifies that decision." Naturally, not everyone sees it that way and we'll return to the contrary arguments later. Next, performance rate which Benson defines as the average rate the plant runs at, divided by the best it has ever run over a period. The fiddle factor is that lots of people divide it by the rate it was designed to run at which can be less or more than it actually achieves. "The argument is if you run it for a month at the [highest] rate, why can't you run it forever at that rate?" Finally quality – the percentage of the product that's right first time. "No messing around reblending or reprocessing or adjusting," he exclaims. "It's what comes right off the line." After this, basic maths encounters Benson's uncompromising realism: "Say your total works out at 40% and world class is 80%. Then theoretically, you could make twice as much out of that plant as you currently do. That's a huge amount of money. Then comes the argument 'Well, even if we could make it, we couldn't sell it'. The answer to that is, you have too many plants." He recalls a visit to a well-known chemical company with the simplest business philosophy: the plant's job was to make as much product as possible as cheaply as it could. Sales' task was to sell it and the two should not be confused. He thinks that's right. The 'hidden plant' is the difference between how much the plant actually makes and how much it could make if it ran at world-class levels. Dividing world-class OEE by actual OEE and multiplying it by gross product margin shows how much extra money could be made out of that asset. He admits it's a pretty broad brush stroke – it leaves out a potential need for extras like raw materials and more labour – but it's still the best indication. Clearly, this view doesn't find universal favour. Malcolm Jones, founder of lean TPM specialist Productivity Europe, thinks it's a flawed theory in an economic climate where supply generally exceeds demand: "It's no use saying 'we can produce it so you should sell it'. What we need to do is to produce what can be sold at the right cost. The capacity opportunity is only an opportunity if there is a market." Instead, OEE should be used to ensure production at the right quality, cost and delivery for existing and potential market opportunities, identified by sales and marketing. Notwithstanding this argument, world class in something capital-intensive like a large continuous petrochemical or mineral processing plant is around 98%. Batch manufacturing is up at about 92%. The best pharmaceutical packing line Benson has ever come across during the BFA judging process was 94%. He recalls other plants in the same sector which could, in his opinion, get three times as much out of that plant as they are currently getting: "These are huge numbers – that's why it's such an important measure. "When I started doing BFA in 1995, probably only one or two out of fourteen [shortlisted sites] would mention OEE. Now it's around eight. When I used to mention OEE in talks, people looked at me as if I was silly. Now virtually everyone knows about it." Even if they don't know how to calculate it? "I would say it is better to measure it incorrectly than never to measure it at all. You are giving some thought to the three factors: the rate you run at, the availability and the quality. And even if you measure it incorrectly but it is improving, at least you are going in the right direction. You can always make it more rigorous as you progress." Not everyone subscribes to Benson's orthodoxy here either. Ask Malcolm Jones what OEE is for and his answer is succinct: "Improvement." "A customer has no interest in your OEE – that is an internal measure which relates to your efficiency and costs," he continues. "The customer is far more interested in a measure like on time in full (OTIF) ie, did I get my order?" He believes that running a manufacturing business on arbitrary efficiency measures rather than a customer satisfaction measure is a recipe for disaster. "The best use of an OEE target such as 85% is to recognise that if you are reaching that level and the customer is still not getting his orders on time, then you may have a capacity constraint." In Jones' view no factory's capacity is perfectly balanced so, in effect, you are only measuring the OEE of the constraining equipment rather than the overall plant. "OEE can tell you how well you are using that constraint. Is it a real capacity constraint, in which case you may need to invest, or is it a constraint because it is being run inefficiently, in which case the improvement cycle kicks in? The point about the improvement cycle is that OEE, through loss analysis, allows you to identify the kind of loss and the required action and then gives a measure of whether it has improved." He points out that Toyota uses OEE only as a spot measure on a particular machine where there is a capacity or quality problem. "Calculating the OEE of anything other than a discrete machine or automated line is pointless; we have far better measures of the efficiency of a factory or department as a whole." Benson would probably not disagree with Jones' view that that the only reason to measure and analyse anything is to improve it. But to one, OEE is a specific measure for use in equally specific improvement projects; to the other, it is the single most important indication of what you could make if you were up there with the best, showing you the cost of falling short of that standard. Interestingly, fiddle factors are equally unacceptable to both (see box on Dos and Don'ts below). There is more uniting their viewpoints than dividing them: the need for unambiguous information about the root causes of downtime; the value of improving changeover time to facilitate smaller batches and meet customer demand without penalty; the importance of a clear measure that helps prioritise improvements that bring the biggest benefits. Whichever way you cut it, OEE is not a measure you can ignore. But there is no doubt that many businesses bite off more than they can chew. Ian Tindle, director of maintenance improvement consultancy Sora, says: "We see it a lot – people want to improve OEE but they won't deploy the right resources. Or they haven't got the resources to deploy. Many say they want to be world class without even understanding what it means. They want to do everything so they end up doing nothing. It's like the Japanese proverb – a man who tries to catch two rabbits ends up catching none." Tindle is adamant that OEE cannot be the sole responsibility of production or maintenance – it's a company objective. But the strategy can and should be boiled down into measurable elements – like breakdown rate, set up, scrap and so on – and cascaded to the relevant departments. Lean devotees will recognise the advantages of Hoshin Kanri in turning strategy into real jobs for real people. Tindle says that often companies that struggle with OEE aren't working cohesively. They may cut set-up losses by 10% but still not see a similar rise in output because downstream operations haven't raised their own game to capitalise on it. Others fall at the first hurdle because they overcomplicate data collection, relying on complex systems that often use the wrong calculations. "So many think they need sophisticated software to do it for them when all they need to start is a piece of paper and an understanding of what's going on at the production line," Tindle continues. "You cannot argue with actual output – if in doubt, start by observing it being made at first hand. In its simplest form, OEE is simply a ratio of how much you made against how many you should have made. If your calculation contradicts this, then it needs looking at again." To Tindle, however, there is one overriding consideration: "Any effort to address OEE can't just be done through technical measurements and improvements. It's all about the shopfloor understanding of what is happening on a daily basis. People need to know what they are doing it for and how it correlates to the overall company objective. The strategic issue has to come first, otherwise it won't be sustainable." how to improve without cheating Don't:

• Massage your OEE
• Log it to planned maintenance when the machine breaks down
• Do changeovers during planned maintenance or at weekends if not 24/7
• Use an easy performance standard
• Measure the best machine and quote that figure
• Set arbitrary targets and achieve them through the above. It may result in decent OEEs and help you make your bonus, but it won't improve your ability to meet customer demand – or even produce more.
  Do:
  • Use it as one of the major tools in improving performance
  • Measure against customer demand (OTIF or similar)
  • Measure OEE on constraints or problem equipment
  • Set realistic performance standards
  • Analyse losses to identify issues for improvement
  • Use the whole DMAIC improvement cycle: OEE Defines equipment performance, Measures it, Analyses the losses, and delivers Improvement and Controls through
  • improved maintenance routines.
  With thanks to Malcolm Jones, Productivity Europe