Plant maintenance: a new smart era of condition monitoring approaches

Discovering which bearing needs changing to prevent production from grinding to a halt could soon be as easy as receiving a ‘WhatsApp’ message from a friend.

“In 15 years’ time the data will be sucked up into mobile phones and the systems will give an engineer an alert saying ‘these five pumps need a bearing changing’,” says Ian Newby, IS development manager at maintenance services firm AVT Reliability.

He believes the current rate of progression in the industry’s use of technology will see inbuilt sensors on rotating processing equipment sending constant data to a system – which then feeds critical information back to engineers through an app on their smart devices.

In 15 years’ time the data will be sucked up into mobile phones and the systems will give an engineer an alert saying ‘these five pumps need a bearing changing’

Ian Newby, IS development manager at AVT Reliability

All this is a far cry from how condition monitoring equipment started out. Designed to create a profit-boosting middle ground between throwing away good kit too soon and letting it run to catastrophic failure, condition monitoring began as an expensive and somewhat cumbersome art form.

“Traditionally the instruments were all custom-made,” says Newby. “A cheap hand-held condition monitoring system once cost more than £10,000. This would be used on every piece of rotating equipment once a month.

“The smartphone in your pocket is probably 100 times more powerful than a data collector was 10 years ago.” A high value has long been placed on the ability to read the future performance of manufacturing kit.

Going round in circles

“Rotating equipment is a key cornerstone of a modern factory,” says Schaeffler UK service engineer Ian Pledger.

An expired bearing can cause a motor to fail, he explains, causing a pump to stop working and possibly result in a production line shuddering to a stop. “So a key driver for most plants is to keep production going,” says Pledger.

Methods of monitoring the condition of rotating equipment cover the measurement and trending of any number of variables. Although oil content, acoustics and heat are often investigated, vibration analysis is the most commonly used.

“In the hands of an expert this will give you very specific information about why a component is failing,” says Pledger.

“Sensors on the machine measure velocity, acceleration and demodulation. Velocity is used for things associated with shaft-turning speed, so predominately unbalance, misalignment and looseness. Acceleration is very useful for gearboxes; we use demodulation for early-onset bearing defects.”

We are looking at the digitalisation of the production chain to such an extent that the machine becomes autonomous

Ian Pledger, UK service engineer at Schaeffler

He agrees with Newby that advances in computer technology continue to modernise the maintenance industry.

Schaeffler this summer launched its FAG SmartQB system which comprises a sensor unit, a housing with a touch panel and a cable for power and data transmission.

Developed to detect irregularities in electric motors, pumps, fans and their rolling bearings, the system performs automatic fault assessment, enabling users to initiate maintenance work and order replacement parts.

Pledger believes this is only the start of things to come. “We are looking at the digitalisation of the production chain to such an extent that the machine becomes autonomous,” he said. “Smart QB is Industry 4.0 ready – it’s ready to talk to the greater machine. It is ready to interact with ordering systems; that’s not too far away.”

Seeing into the future

The possibilities once monitoring instruments start communicating with other machines are intriguing.

“You could get to the point where machines are operating autonomously in a dark factory, where people are only involved to restock or maintain them,” says Pledger.

Glimpses of the future are already here. AVT Reliability’s Machine Sentry condition monitoring system includes an automatic fault-diagnosis assistant that can be used to help identify potential issues with rotating equipment.

“The fault-diagnosis assistant algorithms have been developed by our team of vibration experts using thousands of hours of vibration data coupled with field experience,” says the company on its website.

“The fault diagnosis assistant acts as an aide to the on-site maintenance team in identifying potential causes for concern.”

The more data you throw in, the better your result can be. We have one massive database and we could look across all the assets of a certain type and use all the data we have

Newby believes this type of technology will improve further. “In 20 years’ time we might see engineers getting readouts saying exactly which bearing needs changing – we are on target for that,” he says.

“The more data you throw in, the better your result can be. We have one massive database and we could look across all the assets of a certain type and use all the data we have,” says Newby. “That could be programmed into software.”

As well as receiving information about problems, mobile devices could be used as part of the condition monitoring kit itself.

AVT already has mobile condition monitoring systems that run on off-the-shelf personal digital assistants (PDAs). The firm has been testing its software on Samsung’s Galaxy Tab Active rugged tablet.

“I see a time when people will do most condition monitoring using their phones,” says Newby.

“The technology is getting there but the biggest issue at the moment is battery life. You won’t last a full day. Tablets are generally a bit better.”

View from afar

Emerson reliability solutions Europe director Jean-Luc Goutagny says companies like his could soon be diagnosing invisible problems with rotating equipment from afar.

“The sensors and instrumentation already exist as does the wireless infrastructure in many plants,” he says.

“Once equipment manufacturers and maintenance service organisations have access to this infrastructure, then remote monitoring and diagnosis will become more widely established.

“Organisations will need to make a choice as to whether or not to outsource this service. Do they want to do diagnostics in-house, subcontract to a company like Emerson, or do they want a mix?”

Meanwhile, thermal imaging is gaining popularity as a condition monitoring method as technology improves and becomes cheaper. “Vibration monitoring is approximately 60-70% of the predictive maintenance technology market,” says Goutagny.

“Other advanced technologies such as oil analysis – to determine that lubricants are clean and whether the oil particles indicate wear – and infrared thermography to detect excessive heat or heat loss, are commonly used alongside vibration monitoring data.”

Once equipment manufacturers and maintenance service organisations have access to this infrastructure, then remote monitoring and diagnosis will become more widely established

Jean-Luc Goutagny, Europe director at Emerson reliability solutions

Pledger believes improvement of condition monitoring effectiveness through technology will reduce downtime in factories.

“You can get high levels of uptime using [modern] techniques if they are used correctly in a mature system where alarm limits have been honed,” he says.

“The smart sensors that come with SmartQB have their own learning function, so they learn the vibration of a machine they are fitted to and create alarms based on these.”

As always, the technology needs to be applied in the right way to get good results. The sensors learn what the sound of a normal vibration is like so they can then identify how that compares to a close-to-failure vibration. As a result, if the machine is already in bad working order when the sensor ‘learns’, it is effectively teaching it the wrong ‘normal’, so it has no chance of identifying the early signs of failure.

Visually aware

Pledger continues: “If a machine is older, then you have to be careful. The main thing to do is talk to the operators of the machine; they can be the source of very valuable information. The technology helps us do what we need to do – but there is still a need for human experience.”

Newby agrees, adding that one of the most important conditional monitoring techniques is visual inspection.

“So there is a benefit to having a manual data collection route as someone is looking at the kit.

“There is also a lot of the industry that will never be fully automated. Companies will need people to change tools.”