Keeping a jet plane in the air is a little more mission critical than keeping a biscuit production line running. But if failure were not an option in the latter as well as the former circumstance, productivity in food production might just improve, observes total productive maintenance (TPM) guru Peter Willmott.

But this change of mindset requires cultural change, which is particularly hard to achieve when the person best placed to drive it the ops manager is running around like a headless chicken to keep his plant running around the clock to keep Tesco & Co happy.

"If the engineering manager is given the job, meanwhile, he then has the PIG (production is God) problem to deal with," says Willmott. While most practitioners now agree that maintenance should belong to operations so that overall equipment effectiveness (OEE) is the shared goal, in practice, this sometimes means that preventative maintenance (PM) drops even further down the priority list.

"You've got a four-hour PM task scheduled for Thursday, you get everything ready and then on Wednesday Tesco increases its order and production tells you to cancel," says Willmott. "But when things go wrong, the 'I operate, you fix' mentality kicks in," he says. "After 20 years of talking about TPM. I'm still banging on about the same things." Whether you are a fan of TPM or reliability centred maintenance (RCM), he says, "you have to get maintenance into the boardroom and you have to get frontline operatives more engaged"

While the judicious use of condition monitoring (CM) tools (see Food Manufacture, October 2009, p11) will pay dividends, he adds, the best CM tools are also the most underutilised: the frontline operators' eyes and ears. "When things go wrong or you start to get wear and tear, they look or sound different. They leak, overheat or make a different noise."

Acronyms aside, if operators were properly trained and incentivised to look for these things and do something about them rather than ignoring them and waiting for an engineer to resolve the problem during a weekly check or when the component eventually fails, think how much more efficient your plant could be, points out Peter Bowen at consultancy MPL.

He adds: "The best plants have technical operators basically front line operators that can conduct first level maintenance tasks. They also recognise that engineering management is a professional activity a management discipline just the same as operations management."

For every 100 unplanned breakdowns or stoppages, claims Willmott, "40 could be eliminated by restoring equipment to standard conditions, 20 through front line operator asset care checks, 25 by applying regular condition monitoring and planned maintenance and 15 by designing out physical weaknesses in the equipment"

Reliability centred maintenance

RCM hasn't been 100% successful in food manufacturing, admits Richard Kelly at consultancy EMS, partly because in the early days, proponents stuck too rigidly to models developed for the aircraft industry, and initiatives were often led by engineering rather than production/operations managers.

But the central tenets of RCM are fundamentally sound, and some leading food manufacturers notably Heinz and Aunt Bessie's, have achieved some impressive results, he says. Heinz has recently increased the availability of a rotary blancher and reduced engineering man hours where it has tried out RCM in parts of its Kitt Green factory in Wigan, while Aunt Bessie's has achieved similar results at its site in Hull.

A key aspect of RCM is the recognition that the timing of component failure is random, and does not conveniently coincide with scheduled production stoppages, says Kelly. And yet many firms still base PM tasks on a simple time-based system (replace part x every 12 weeks) intrusive maintenance that is costly, time consuming and not actually particularly effective at reducing unplanned downtime.

RCM instead encourages firms to mobilise front line operators, says Kelly: "With the minimum of training, 80% of maintenance tasks can be done while the line is running by the operators themselves." Demonstrating what things should look like when they are working optimally either through pictures, arrows, or other visual cues and making key components easier to see and access, is also critical, he says.

RCM also recognises that sometimes the only way to avoid failure is to redesign equipment or how it is operated. "If water getting into bearings is the problem, can you stop the water getting in, and if you can't do that, will a plastic bearing work instead?"

Similarly, instead of arguing over whose fault it is when an operator drops a change part into a piece of kit during a changeover, damages it and delays production, RCM encourages a different approach, says Kelly. "The downtime might be attributed to production, but perhaps the solution lies with engineering can they design a solution so that it is not possible to drop the part into the machine in the first place?"

Implementing RCM at Aunt Bessie's has achieved the primary aim of increasing availability and reducing unplanned downtime, says engineering team leader Lee Tyler. But the wider benefits of improving line operators' understanding of how the plant actually works have been just as significant.

"We show them what a gearbox, fan or motor looks like, what it does and what it should sound like, so they know what to look and listen for." This approach has been so successful that more than 80% of maintenance tasks are now carried out by operators compared to virtually none before RCM was implemented, he says.

Tyler's team also identified more than 50 'design outs' to increase reliability and make equipment easier to monitor or repair in the first place. After all, maintenance without problem solving and development engineering is always going to be a waste of time and money, points out Lauras International Europe partner Simon Spanyol. "At one firm I worked in, we used to use Bepex Hutt co-extruders and over time every component was redesigned and rebuilt from different materials with a result that they needed less maintenance and didn't fail."

Equipment suppliers are getting better at designing kit so it is more accessible and easier to clean, inspect and maintain, with better visual cues and more quick release mechanisms, but if things are going to improve, buyers need to build reliability into their specifications, says Peter Gagg, joint md at consultancy MCP.

As for factoring in sufficient time for PM tasks, factories running five days a week instead of seven are not necessarily any better at this, notes Lauras partner Jeremy Praud. "On a 24/5 operation, you leave everything until the weekend when it costs you more to have engineers and everything has to be done at once, whereas on a 24/7 operation you have to plan ahead more carefully."

Depressingly, he adds, there is often no correlation between the sums spent on maintenance and factory performance, which means a lot of people are clearly not getting much bang for their buck. "Sites that spend the most on maintenance often have the most [unplanned] downtime."

Computerised maintenance management systems

Similarly, firms with a fancy computerised maintenance management system (CMMS) do not always get the payback they expect, says Gagg. "People are often just using them to generate preventative maintenance work orders and record breakdowns, and even then you often find that they are only recording them when a line has stopped for more than 15 or 30 minutes or if a replacement part is required, and they are not learning from the information they record."

When used as an analytical tool to help managers see what is causing the most persistent problems and then allocate resources into problem solving or development engineering, CMMS is invaluable, however, says Andy Neilson, marketing boss at Spidex, the UK and Ireland distributor for Mainsaver enterprise asset management and CMMS software. For compliance purposes, meanwhile, being able to access and interrogate maintenance plans and records in seconds is also critical.

But mostly, a good system used correctly can stop every day turning into groundhog day, says Idhammar operations director Alan France. Take the following example: A call comes through that machine A is down and it's something to do with the glue unit; an engineer is despatched, discovers a nozzle is blocked, which requires a certain spanner to remove which he hasn't got with him. So he goes back to get it, returns, realises a new nozzle is in fact needed, hunts around to find one, and can't, so orders a new 10-pack. When the nozzle blocks again a month later, a different engineer is on duty and doesn't know that the other engineer ordered in a new pack or where he left the spares, says France, "and the same thing happens all over again"

Replay this scenario with an effective CMMS, he says, and history might just stop repeating itself. The call comes through, the engineer looks up the machine in the CMMS using a hand-held PDA to see if there's a history of problems with the unit, when it was last maintained and if there are any notes. He then checks to see if more nozzles are in stock via the PDA, which tells him exactly where the spares are located.

Finally, he goes to the machine with the correct spanner, fits the new nozzle and makes a note in the CMMS flagging up the nozzle as a recurring issue so that new work orders are scheduled to ensure it is replaced at more regular intervals or monitored more closely. FM