Predictive maintenance - as a concept – has existed for a long time.
What’s new, however, is the availability of data and the ability to analyse it practically, quickly and cost-effectively. The use of IIoT technologies will not only predict which failures could occur, but when that is likely to be.
Imagine an unexpected issue arises in your production environment – an abnormally vibrating pump, for example.
Ruminating over the possible causes (and outcomes if it fails) can result in many sleepless nights for a plant manager or maintenance engineer.
Why? Because if the pump stops working, it could lead to costly unplanned maintenance and production downtime.
“In some sectors, such as water treatment, an entire plant can be taken offline by the failure of a single part,” says Richard Jeffers, founder of RS Industria, “which could be a potential cost of around £20,000 per hour.”
So how do you get under the hood of this problem to prevent such a costly failure? The answer lies in predictive maintenance, which involves monitoring the condition of machinery to identify when a part will fail.
This not only reduces surprise failures and the associated downtime it takes for an engineer to fix the problem, but it also removes the need for scheduled maintenance and the routine replacement of parts that may, in fact, be perfectly healthy.
How has predictive maintenance evolved?
“The core concepts around maintenance were developed out of the aerospace industry in the 70s and 80s,” says Jeffers. “So, we’ve known how to do predictive maintenance based on conditioning monitoring and the analysis of data for over 40 years.”
This early condition-based maintenance involved monitoring and analysing the equipment’s physical condition and the frequency of its use.
Using this data, adjusted maintenance processes and inspection cycles could be formed. Of course, in the 70s and 80s, this would have been done manually by data experts who would develop custom models for each type of machine, with any changing patterns in machine behaviour indicating that trouble could be on the horizon. This enabled plants to be reactive to maintenance needs.
The difference today is that IIoT technologies use advanced analytics to dissect machine data automatically - without any human intervention - and this capability can be deployed at scale.
“The key is that IIoT can unlock data to provide insights into the overall effectiveness of the equipment and predict defects long before a human can,” says Jeffers. “We now have the tools to easily extract this data, making it available at the right cost.”
As a result, IIoT provides an even better understanding of the ongoing health of machines, enabling action to be taken before equipment fails and even provide insight into when that is likely to be. From a reactive and proactive approach to maintenance, we are now truly moving into a predictive and preventative approach.
How does predictive maintenance work?
Crucial to predictive maintenance is the extraction and collection of data from machines, which involves investment into an IIoT platform like RS Industria.
“By itself, data may be interesting, but it’s how you get data out of the machines, enable users to easily interact with it, then analyse that data to provide insights that’s key to transforming your business,” says Jeffers.
It’s often thought that an IIoT platform has to be considered at enterprise level, but predictive maintenance can be applied at just machine level. “RS Industria is modular, meaning that users can start small with just the one critical problem that they want to solve,” adds Jeffers.
RS Industria’s modular solution allows you to start your journey with critical assets and expand as you gain value.
We understand that you need to address your most pertinent pains first – and our rapid approach means we can alleviate those pains in weeks, not months or years. Whatever is keeping you awake at night – be it high energy costs, unplanned downtime, high emissions, high waste or production yield – we can focus on extracting the asset data that will bring you the quickest return on investment.
More assets and production lines can be added if and when required (with known, predictable costs), allowing you to retain complete control.
Furthermore, we’re brand agnostic and can connect a range of PLCs, automation systems and sensors to our solution – as well as saving time for busy maintenance and production staff by managing the entire connection and configuration process on your behalf.
How does machine learning assist predictive maintenance?
As well as monitoring the live data coming in, predictive analytics can also be applied to that data.
For this application, RS Industria has partnered with Senseye, a leading provider of predictive maintenance software solutions.
Senseye makes it possible to apply predictive maintenance in a novel, resource-efficient manner - with minimal training - by utilising rules-based alerting and machine learning.
The benefits include:
- Reduction of downtime by 50%.
- Reduction of waste by 40%.
- Early warnings to enable a rapid response to maintenance issues.
- Asset problems that are highlighted ahead of time so your maintenance teams can intervene.
In simple terms, ML (machine learning) uses algorithms to find patterns in data to predict future events.
By combining real-time data with historical trends and variables such as current environmental factors, ML can make decisions on when action needs to be taken. Maintenance teams can then carry out this action without actually understanding these algorithms.
“Operators can prime the system upfront with helpful information, such as data recorded in the run-up to previous failures,” explains Robert Russell, CTO at Senseye. “But the algorithms are designed to start from day one and will provide useful insights soon after. They will get more accurate as they learn from your machine’s and maintenance team’s behaviour.”
Plant managers no longer need to suffer sleepless nights, worrying about a part failing. They now have the insights at their fingertips to know when that is likely to be, and step in to prevent it.
“It is a solution that is absolutely practical right here, right now,” concludes Jeffers. “It’s an enormous leap forward for any plant or factory that has struggled with unplanned downtime.”
RS Industria unlocks the valuable data hidden in your factory - so you can increase uptime, reduce losses, and create new operating insight.
Our solution is the key to improving the reliability, sustainability, and performance of your manufacturing assets. Better yet, it’s simple, fast, and affordable to implement.
Ready to learn more?
Read on to ‘Case Study: Reducing Energy Usage Through Insight at Kerry’.