Connected technology with a purpose


SKF’s Rotating Equipment Performance Centre.

Farrukh Yaqub, head of rotating equipment performance at SKF Australia, considers the importance of keeping purpose in mind when implementing digital mining solutions.

Digital tools for equipment monitoring are getting smarter, more powerful and more cost-effective. But as mine operators implement these new technologies, they should always keep their purpose in mind.

Predictive maintenance has traditionally been used by early adopters with an understanding of how strategic maintenance management can contribute to profitability.

High costs and complexity once limited predictive maintenance to the most critical, industrial assets. Today, exciting changes are afoot.

“New developments, including mobile computing, wireless communications, Internet of Things (IoT) and artificial intelligence (AI), make it feasible to add smart monitoring capabilities to a much wider range of machines,” Farrukh Yaqub, head of rotating equipment performance at SKF Australia, said.

“In such a fast-changing and dynamic environment, it is easy to get carried along by the hype.”

With technology companies and service providers offering to wire up anything and everything, the key question for mine operators is no longer whether they can monitor their assets, but how to manage the increasing volumes of data.

To find purpose, equipment operators need to begin with a clear vision of their end goals.

What are they trying to achieve with their overall maintenance and reliability program? How could digital technologies contribute to those aims?

In SKF’s conversations with miners, the company has heard several different answers to that latter question.

“Often they have strong financial goals, such as maximising the return on capital investments or minimising the total cost of ownership (TCO) of their assets,” Yaqub said. “Improved safety is another universal priority in demanding and hazardous mining environments.

“Resources companies are also prioritising actions that improve their operational sustainability, as part of the license to operate and to make it easier to attract and retain a skilled and motivated workforce.”

SKF’s IMx-8 technology is advancing smart monitoring capabilities.


Setting out the goals of a potential machine-monitoring and predictive maintenance program provides a framework that helps owners prioritise their actions and implement the most appropriate solutions.

Mining equipment will always be exposed to wear and the risk of damage, but by identifying the assets and failure modes that have the largest impact on their reliability goals, operators can identify technologies that help them detect, predict and prevent those failures. 

Once mining companies have defined the objectives of their monitoring and predictive maintenance project at the asset level, it is time to identify the right solution. That solution will typically comprise several connected hardware, software and human elements.

SKF finds it useful to break the solution down into four distinct parts: connect, detect, inform and improve.


‘Connect’ is about the equipment and infrastructure needed to measure and record reliability-related data. For bearings and rotating machines, that usually involves vibration sensors, since vibration is still the best way to spot early problem signs.

However, condition-monitoring systems are increasingly combining multiple data types, such as process data and production data from machine control systems or lubrication oil temperature and condition information from dedicated sensors.

Combined with handheld devices, sensors can be used to collect data periodically during walkarounds by maintenance staff. Or they can be permanently fixed to the machine, transmitting data via wired or wireless connections at an appropriate rate.

“The best approach depends on the type of asset being monitored and its role in the process,” Yaqub said. “Traditionally, permanent sensors and wired connections were expensive to install and were only used on the most critical assets.

“Balance of plan assets, such as pumps and fans, could be effectively monitored using handheld devices.

“Attitudes are beginning to change, however. This is being driven by the development of lower-cost wireless sensors that are quick and easy to install.”

Switching from handheld devices to wireless sensors frees up maintenance staff for more value-adding work and has come with safety benefits, reducing the need for personnel to approach running machinery for data collection.

Continuous monitoring can also allow for quicker problem identification than the traditional two- or three-week manual data-collection cycle, helping to reduce environmental concerns and save the planet’s resources.


‘Detect’ entails the analytical techniques used to identify anomalies in machine data and diagnose potential faults. This is a highly specialised task requiring a combination of sophisticated algorithms and human expertise.

Difficulty accessing specialist personnel can be a critical bottleneck for predictive maintenance programs, especially for miners operating in remote and inaccessible locations.

Today, the need for on-site expertise has been lessened by the development of centralised remote monitoring facilities, which use cloud technologies to share data from multiple sites and assets from a single hub. Specialist staff then analyse the data and diagnose problems.

“Some large organisations choose to staff and run such facilities themselves, but this can also be outsourced to specialist providers,” Yaqub said. “It is a service SKF provides for its customers worldwide through its network of remote diagnostics centres (RDC).

“There is also an ongoing drive to automate more anomaly detection and problem diagnosis using AI technologies.

“AI systems are already showing tremendous promise in machine monitoring applications, both by reducing the workload placed upon human experts and by spotting subtle signals people might miss.”

SKF’s head of rotating equipment performance Farrukh Yaqub.


‘Inform’ is about translating the anomalies detected by the remote monitoring system into actionable information for the end user. For example, if a problem is detected in a bearing, the user knows they will need to inspect the machine.

What the user really wants to know is whether they should stop immediately or whether it will be safe to keep the machine running until the next scheduled maintenance intervention.

SKF’s experience suggests the ‘inform’ phase works best when it involves close collaboration between condition-monitoring and data analysis experts and the miner’s own operations and maintenance teams.


The need for collaboration is even more pronounced in ‘improve’, which is often key to the most significant performance and value improvements.

It is about the robust application of traditional reliability improvement tools to prevent the reoccurrence of failures identified by the predictive maintenance system.

SKF harnesses ‘improve’ by working with its customers to identify changes in bearing selection or lubrication strategy that can prolong the life of equipment and extend the mean time between critical asset failures.

Case study

A horizontal grinding mill was installed at an Australian mineral processing site, presenting some significant challenges.

The remote location of the site meant access to specialist personnel was not always available. In the past, the customer had experienced a situation where a critical failure was picked up too late, leading to a major unplanned shutdown.

A second challenge related to the nature of the mill operations. The asset rotated at low speeds of around 10 revolutions per minute (rpm).

This necessitated the use of highly sophisticated vibration analysis algorithms to extract valid data on the condition of bearings and gearbox elements in an extremely noisy environment.

SKF provided a mill-monitoring system that attended to the customer’s unique environment and situation.

“SKF monitored the mill on behalf of the customer, installing the necessary sensors and communication equipment and conducting analysis remotely from one of its RDCs,” Yaqub said.

“After only three months of continuous monitoring, the RDC identified an anomalous signal and was able to pinpoint the problem to a particular pinion gear within the machine.

“In this case, timely action to the problem saved the customer enough in avoided downtime to cover the cost of the remote monitoring system.”

New developments make it feasible to add smart monitoring capabilities in a much wider range of machines, but it is easy to be carried away in the hype. Break down the solution into four parts: connect, detect, inform and improve.

The key question that follows is: what is trying to be achieved with the overall maintenance and reliability program?

Once this is answered, the solution can be implemented. 

This feature appeared in the April issue of Australian Mining.

To keep up to date with Australian Mining, subscribe to our free email newsletters delivered straight to your inbox. Click here.