This website uses cookies primarily for visitor analytics. Certain pages will ask you to fill in contact details to receive additional information. On these pages you have the option of having the site log your details for future visits. Indicating you want the site to remember your details will place a cookie on your device. To view our full cookie policy, please click here. You can also view it at any time by going to our Contact Us page.

Predicting the future of maintenance

25 June 2017

Meeting the challenges facing the food processing sector today requires an increase in machine availability and a reduction in unscheduled downtimes and so it is important to look at techniques that can help to manage maintenance and maximise production reliability. Suzanne Gill finds out how advancing technologies and digitalisation of the plant floor might affect maintenance strategies for food processors.

Reactive maintenance is the worst possible scenario – fixing problems after they have caused unscheduled downtime. The failure could be something as simple as the ink running out in a labeller, but the resulting stoppage can cause significant impacts on costs, productivity and reputation. 

Preventative maintenance is an improvement. Impending problems can be tackled during scheduled downtime. However, this often relies on products being replaced based on statistical lifecycle data. So you might be switching out components that still have years of useful life. On the other hand unexpected failures could still occur.

Condition monitoring, in contrast, offers a predictive approach to plant maintenance. Predictive maintenance is built on the real-time monitoring of actual machine parameters to help eliminate downtime and it is an important step in maximising productivity.

Jeremy Shinton, product manager business solutions at Mitsubishi Electric, highlighted a new technology from Mitsubishi Electric that is designed to aid predictive maintenance. The Smart Condition Monitoring (SCM) solution combines FAG SmartCheck sensors from Schaeffler with Mitsubishi Electric control technologies, which together can monitor a full range of machine parameters. 

“The SCM analysis provides detailed diagnostics, and provides maintenance staff with precise error identification,” said Shinton. “It provides recommendations on what remedial actions should be taken, with text messages presented to personnel. Once set up, the system provides 24/7 monitoring of assets, with functions including bearing defect detection, imbalance detection, misalignment detection, lack of lubricant, temperature measurement, cavitation detection, phase failure recognition and resonance frequency detection.”

Driving change
According to Steve Sands, head of product management at Festo, increasing digitalisation is driving the change in approach to maintenance. “One outcome of this is the evolution of condition monitoring, which enables more predictive maintenance based on data about the product, the environment, wear and tear and indication that faults are developing. The ability to anticipate issues means that maintenance outages can be better timed and the causes of unplanned stoppages can be avoided altogether,” he said.
Advances in automation have the potential to further evolve the maintenance approach. “Industry 4.0 developments, for example, will see machine learning becoming more common, with embedded intelligence allowing Artificial Intelligence (AI) algorithms to detect data patterns and to propose corrective measures to the maintenance team,” continued Sands.

He explained how augmented reality is now also breaking out from research and ‘exotic’ applications into mainstream production. “Blending images and design data with the real situation and providing the possibility to look inside will aid understanding and fault resolution. The concept of a Digital Twin – a CAD/CAE -generated model that enables designers to simulate how a machine will function – is clearly already possible. Industry 4.0 will see this model linked in real time to its physical counterpart, enabling the virtual model to emulate the machine throughout its operational life. Information about smart components will be held within the model and continuously updated. 

“The Digital Twin of the future will capture information about modifications as they are made, eliminating the need to revise technical manuals as variations to the original design are introduced,” said Sands.

Clearly, this level of automation will have significant implications for maintenance teams. They will be able to take a fully proactive approach to maintenance, using handheld technology to analyse condition data and interrogate alerts from anywhere within the factory. Automatic ordering of tools and parts will increase efficiency by eliminating journeys between the stockroom and the machines requiring attention. Ultimately, Industry 4.0 adoption will see the role of maintenance elevated as an integral part of the food production process, rather than the symptom of unwelcome breakdowns.

Also commenting on how technology advances are making predictive maintenance strategies possible, Keith Thornhill, business manager food & Beverage at Siemens UK and Ireland, said: “Advanced diagnostic and sensing technologies include ultrasound detectors, thermography, vibration and oil analysis, to identify failures at the earliest possible stage. One example of this in practice would be the capture of thermal radiation images of equipment and components, which can be successfully used to find potential issues on a wide variety of equipment types.” The images can be used to discover hot areas of components (due to excessive friction, worn or loose parts, poorly connected or deteriorating wiring, overcurrent conditions, lack of cooling and unbalanced loading, for example) that are indicators of potential failures.

“While this data can be used to identify failure causes, it is vital that food manufacturers also have a robust plan in place to successfully extrapolate this data, using machine learning to analyse this data, identifying patterns, so, should machine failure occur, its trace is captured. This means that in the future should the same trace be identified, measures can be put into place to prevent failure and subsequent downtime which incur costs to a business,” concluded Thornhill.

A knowledge gap?
So, while advances in technology are making predictive maintenance a real possibility, Martin Walder, vice president Industry UK & Ireland at Schneider Electric, questions whether there is sufficient knowledge in the food factory to utilise it, due to a gradual reduction in the number of production engineers working in the food manufacturing  environment in recent years. Walder said: “The utopian predictive maintenance scenario would be to have the ability to predict any failure, in advance. It would not then be necessary to hold any spares and plants could run with fewer maintenance staff, with activities being scheduled at the most suitable times. While I don’t think that this will become reality for some time, it is already possible to get a good indication of failure from a lot of devices that are already working on the plant floor today. The technology is already here to allow you to look at the performance profile of a drive, for example, cycle after cycle after cycle, and this allows you to detect when it was drifting from normal which would indicate a potential problem.

 “However, this capability is often not utilised because many would struggle to justify the cost of getting an outside engineering company to create a solution to allow the data to be gathered and analysed to turn it into actionable information,” continued Walder. He advises that together clear ROI plans is key to getting a predictive maintenance system approved. “It is always important to be able to demonstrate that it is possible to get a clear return on investment for a project,” he said. 

“The good news is that virtually all automation technology being sold today has IoT capability included as standard. It is just a question of gaining an understanding of how to use it,” continued Walder. He believes that the capability for predictive maintenance is most likely to be supplied by machine builders and advices that you as ask whether they are offering IoT enabled technologies when discussing a new line or a system upgrade. “Even if you are not buying a machine with a service, it is important to specify connected devices, concludes Walder. “The cost is now so similar to a non-connected devices that it does not make sense to not specify it. Even if you are not going to use it immediately you will be futureproofing your plant.” 

Contact Details and Archive...

Print this page | E-mail this page