Check, act, plan, do

Most plants and factories have some form of CMMS (computerised maintenance management system) – at the very least spreadsheets – but few use them well. So says Dennis McCarthy, director of lean manufacturing consultancy DAK Consulting. And he believes that, in large part, that's down to the role of maintenance and how it's perceived in the organisation. "If maintenance technicians are seen as just there to fix breakdowns and stabilise the plant, they won't be highly valued. So CMMSs won't be used to anything like their full capability – probably for little more than routes and controlling the cost of spares," he explains. "If, on the other hand, engineering is about optimising the plant, then they get more recognition, because not only are they responsible for keeping everything in good condition but also for helping to drive CI [continuous improvement], and hence efficiency and profitability." And we might add: product quality and throughput, as well as energy efficiency. This, in turn, makes their CMMSs critical to sustaining the business, which means they're on the management radar, so set up better, kept up to date and integral to the whole CI endeavour – and you have a virtuous circle of improvement. But that doesn't happen by itself. As McCarthy puts it: "CMMS is a tool to support good management. However, if you haven't got that, then it becomes a tool to support bad management." Then you're into fire-fighting just to keep the plant running, plans go by the board and the system falls into disrepute. That said, it's also worth noting that plenty of factory managers are pushed into accepting sub-optimal CMMSs on the mistaken premise that you don't need systems to run RCM (reliability centred maintenance) or TPM (total productive maintenance). Well, maybe you don't, but try running a world-class factory without systems to support the efficient engineering that is essential to properly functioning plant and equipment with long MTBF (mean time between failure) stats. So what does 'good' look like, when it comes to CMMSs? McCarthy suggests observing process plants that must be kept up and running for weeks, months or even years. "The focus for maintenance there is always on optimising plant life and uptime, so CMMSs are applied very well. They're used to document best practices – those that result in quick and efficient service, repair or replacement without installation errors – as well as to feed back equipment condition, track interventions and modifications, and check that plant is doing what operators expect, from a performance perspective." Clearly, the message is that good, well-applied CMMSs are not just set up to plan and schedule preventive maintenance, and the associated spares, purchasing and supplier management. They're also configured to be forward looking, focusing on monitoring and managing plant improvements. A CMMS designed only to support planned maintenance isn't going to cut the mustard. McCarthy again: "With planned maintenance, the best you can hope for is that this year will be no worse than last. The role of maintenance should be to go past that, so you need processes and systems that do,too." Indeed, in some cases, good CMMSs extend beyond maintenance to helping plants gear up for flexibility and responsiveness, in terms of anything from customer service to new product introduction. And that's where EAM (enterprise asset management) systems come in. Jean-Benoit Nonque, regional vice-president for EAM systems at software giant Infor, describes these as essentially adding integration into production and business systems. "With EAM there is more flexibility in terms of information and coverage," he explains. "So, for example, operations and engineering can run mobile systems, see that parts are available for preventive maintenance but could also make sure that spares inventories are optimised so that costs are minimised, parts don't become obsolete before they're used and supplier numbers are rationalised." And he adds that, with production and maintenance talking at the system level, not only can interventions be planned better, but manufacturing quality issues can be tackled at source. Devil in the detail But the devil is in the detail. Good CMMS/EAM is not just about setting up a hierarchy of functional locations and associated equipment with some workflow. For CMMSs to be effective, they need to be fit for purpose – and that inevitably involves a trade-off around the scale of data. "The more detail you build in, the more you have to maintain," warns McCarthy. "So you have to ask, 'where's the value add?'. We worked on one project with BT where the goal was to get down to the big parts that wear in three clicks. So that dictated our detail." For him, it's best to build the database so it allows "maximum intervention with minimum difficulty". It's about getting enough information to be useful, in terms of both maintenance and improvement, but no more. For example, it makes sense to capture operational costs, not just criticality, to major assets. Why? Because, the rate of rising costs provides a useful indication of when assets need overhaul or replacement. However, for many plants, there's little point in recording routine stuff, such as bearing temperatures, if they're in normal range. That's one view. Another, as expressed by David Manning-Ohren, condition monitoring business unit manager with MRO (maintenance repair and overhaul) specialist Eriks, is that CMMSs are necessary but not sufficient. If you're interested in preventing recurring failures and/or minimising asset repair and replacement costs and timescales, he says, you may need more detail than you think. And certainly more than some CMMSs cover. That's why Eriks typically uses its own 'repairable asset management system'. He explains that, when it comes to electric motors, for example, you need the facility to search on data such as foot- or frame-mounting and power output, but also failure modes. That's because there might well be something similar on a gearbox or pump that could enable rapid replacement in the event of failure. Equally, if an upgrade is required – say, from a roller to a ball bearing – you need to identify all items of similar construction on comparable duties to find those at risk. Then again, there may be similar failure modes at a particular location – maybe due to misalignment, poor set-up or contamination – and there's little point in replacing like with like, without remedying the situation. "At Ford Motor Company, for example, back in the late 1990s, we were able to use our system to identify a lot of brush gear on servo motors, which was being attacked by lubricants used for cutting. We replaced all the brushed motors with brushless to solve the problem," says Manning-Ohren. Spares listings on CMMSs, he says, are typically just not detailed enough for this sort of activity. "We've just won a contract with Tata Steel, which uses SAP's EAM for purchasing and inventory management," states Manning-Ohren. "We're rewriting our repairable asset management software to integrate with SAP so that we can share information and track items with the data we need for this fixed-price MRO contract." His advice: implementing a repairable asset programme will save enough money on repairs to pay for the condition monitoring that, in turn, reduces the number and extent of failures on plant. "That's just the repairs, not the big stuff, such as the downtime itself," he adds.