Featured

What does a successful predictive maintenance program look like?

What does a successful predictive maintenance program look like?

Freddie Coertze wants to make a distinction between predictive maintenance and condition monitoring.

“If I were to summarise the key differences it would be to say that condition monitoring is a reactive approach to maintenance management that focuses on identifying problems as they occur, while predictive maintenance is a proactive approach that uses data analysis to predict when maintenance will be needed,” says the national IoT business manager for ifm Australia.

“Both approaches can be valuable, but predictive maintenance is defined by the word ‘predict’. It has the potential to provide greater benefits by reducing downtime and minimising maintenance costs.”

According to Coertze, there are three components that can ensure the success of a predictive maintenance program: An Internet of Things (IoT) platform, condition-monitoring hardware and predictive formulas provided by artificial intelligence.

“Importantly, to have visibility of assets, you need the connectivity and integration that digitalisation provides,” Coertze remarks.

“To help our customers simplify this integration, ifm created moneo – a self-service software platform that acts as a middleware to existing systems such as SCADA.”

The moneo solution has an inbuilt DataScience Toolbox that enables operators and engineers to leverage their knowledge of equipment with the benefits of AI-assisted predictive analytics and machine learning.

“Successful predictive maintenance is about detecting anomalies to machine performance early and automating the actions surrounding that deviation so that not only are notifications sent, but actions are prescribed to fix the issue,” explains Coertze.

“It’s having full integration from the shop floor to the top floor.”

Currently, Coertze says that while condition monitoring is widely used, there is a disconnect between the data collected by sensors and the workflow.

“Many industrial businesses still rely on manual processes in terms of how their sensor data is analysed and maintenance is performed. Many businesses are still using a preventative or scheduled maintenance approach – they might look at vibration on componentry but it’s not going to give them the insight or ability to solve real equipment issues in advance,” he explains.

“I am of the opinion that this type of condition monitoring – where it is applied within a manual context of action – provides information too late to make a significant difference.”

The moneo solution supports successful predictive maintenance by providing the platform, the sensor hardware, and the software that supports AI-assisted predictive formulas.

“To explain simply how it works, we would first connect the sensors to the equipment, then we would collect data from the sensors and the moneo software will identify baselines and set parameters and limits as to how the assets should be performing,” he says.

“If the system detects an anomaly, a ticket will be sent to staff to investigate with a prescribed action to first check and then rectify the problem. This same predictive maintenance workflow is then applied to everything within the operation or plant.”

There are many benefits that come with implementing a successful predictive maintenance – namely better longevity of equipment, improved productivity onsite, and improved efficiency, particularly in the use of energy.

As Coertze concludes, predictive maintenance will do more than prevent downtime, it will help keep equipment running optimally at all times.

“Condition monitoring with vibration analysis is simply not enough – by the time vibration has started, it’s often already too late to intervene and save the machine. To protect your assets, you need to predict. That’s why having a predictive maintenance program is important and moneo can ensure that this program is implemented with success.”

Send this to a friend