It all adds up

Scratch your typical maintenance manager, says Alan France, and you'll find someone with a firm eye on the factory's breakdown log. Which, he adds, makes perfect sense: if vital production equipment is breaking down, then output levels will obviously suffer. But the astute maintenance manager will also keep an eye on another source of factory floor information, says France, the operations director at specialist manufacturing software firm Idhammar Systems. The information in question? The factory's overall equipment effectiveness (OEE) data, which tracks the machine-by-machine impact on production output of such things as yield losses, set-up times, slow running and minor stoppages. "Look at the breakdown log, and you'll see those events where production stopped due to machine failure," says France. "But look at the OEE data, and you'll potentially see much bigger sources of production loss, spread over a number of causal factors with a potential maintenance dimension – but which get lumped together as 'normal running'." Best-case scenario And the logic isn't difficult to understand. OEE takes as its starting point the theoretical 'best case' output level that a machine is capable of – running at rated speeds, with no set-ups, no quality losses, no breakdowns, and no minor stoppages. OEE losses, then, are anything that detracts from this best case. To some maintenance managers, notes France, such losses are the day to day reality of factory floor life. An assumption of no set-ups – and therefore no lost output due to set-up time – is simply unrealistic, they assert. Likewise, they add, it's difficult to argue with quality-related losses and slow-running on 'difficult' jobs: they might be problems, in short, but they're not maintenance problems. Yet such thinking is mistaken, insists France. Such losses do have a maintenance dimension – and possibly a very significant one. And until the maintenance function engages with the problems in question, the extent of that maintenance dimension simply isn't known. "If too much time is being lost on set-ups, there might be something that the maintenance function can do to reduce the time – classic single minute exchange of die (SMED)-style engineering techniques," he notes. "Likewise, quality-related losses or slow-running might have simple fixes, due to settings drifting out of tolerance, or excessive wear on tooling. And until engineering take a look at the issue, production losses will carry on occurring." Daily reviews In short, in France's eyes, there's a lot to be said for simply making time in every factory's regular morning review meeting for a look at the latest OEE data, and involving engineering in the resulting deliberations. "Classically, OEE is used in the context of periodic improvement activities, or logged on a daily basis until some sort of action point trigger is reached," he notes. "But that means there's an awful lot of OEE data capture that is simply wasted, because it's not being used to trap maintenance issues, as well as Six Sigma-type improvement activities. They're quite separate initiatives, and there's a place for both." That said, adds France, care needs to be taken when responding to OEE data. Overly-simplistic OEE analyses – ones that are perhaps chart- or spreadsheet-based – can point maintenance managers at issues that turn out to be symptoms, not causes. "We've seen instances where stripping down a supposedly breakdown-prone unit and servicing it would have been a waste of effort," he notes. "Simplistic charts and spreadsheets don't really provide a 'drill-down' capability, and before committing expensive engineering resource it's best to establish if you're looking at a cause or a symptom. A unit might be failing, but failing due to out-of-specification materials, or operator error." Hence the attraction of OEE systems. As well as producing the sort of charts that production managers like to display, they also provide enough inbuilt intelligence and drill-down capability to help identify real maintenance issues, not phantom ones. "Remember the words of management guru Peter Drucker," he says. "'Efficiency is doing things right; effectiveness is doing the right things.' And knowing exactly how well the plant is performing will help to drive maintenance functions to do 'the right things'. Put like that, it seems obvious that maintenance should take an interest in OEE data."