Predictive maintenance dodges downtime

Advanced wireless sensors are being developed to accompany automated vehicles in next-generation artificial intelligence-operated heavy machinery in mining and other sectors, to help predict malfunctions before they arise.

With a fleet of futuristic self-driving equipment being set up to operate standalone, a system for maintenance of these vehicles is necessary to circumvent untimely failures.

Typically, standard preventative maintenance is applied, which is just routine repair to help keep equipment up and running, preventing any unplanned downtime and expensive costs from unanticipated equipment failure.

Maintenance technicians would rely on preventative maintenance schedules provided by the equipment's manufacturer, including regularly replacing machine components on a suggested timeline.

These, however, are only estimates of when the machine would require service, and the actual use of the equipment can greatly affect the reliability of those estimates.

For example, if bearings wear prematurely or a motor overheats, a machine may require service sooner than projected. Furthermore, if a problem goes undetected for too long, the issue could escalate to further damage and lead to costly unplanned downtime.

Yet, with new technologies, sensors for monitoring the condition of machinery has become advanced enough to allow a new approach to upkeep – predictive maintenance

Predictive maintenance

By understanding the underlying "health" of a hard-working earth moving machine, a mine operator can determine if a corrective action is required before detrimental failure or damage occurs.

This predictive maintenance relies on sensors to measure various machine parameters such as vibration, temperature and ultrasound.

Without the experienced hand of a human being behind the wheel to detect irregularities, the auxiliary system of sensors alongside AI are now being used in tandem to ping vital status information back to the central control station.

Predictive maintenance is only possible with 'condition monitoring' due to these sensors. This is provided by tracking the performance of crucial machine components to show projected issues based on aggregated data.

This allows technicians to more accurately anticipate when repairs will be needed based on real-time data from the vehicles themselves.

Of the sensors, monitoring for vibration is a crucial part in determining the health of machinery, as a rattle in the hood is never good.

Machine vibration is often caused by imbalanced, misaligned or worn parts. As vibration increases, so too can the damage to the equipment.

Motors, pumps, compressors, fans, blowers and gearboxes are some of the major components that are monitored so that problems can be detected before they become severe.

Aside from monitoring vibration, thermal sensors are used as well, as heat is often an early indicator of degradation. And as cold spots often suggest blown fuses or failed capacitors, identifying abnormal thermal patterns before a failure can help extend a machines life.

Thermal imagers - also called infrared cameras - help maintenance, reliability and operations professionals quickly identify hot spots and cold spots that could indicate potential problems, helping them avoid unexpected downtime and equipment damage.

Maintenance technicians use infrared cameras to inspect entire areas or a single piece of equipment at a specific time. As anomalies are identified, the technician can zero in on irregular areas to determine the magnitude of the problem.

Yet, maintenance personnel would rather find an impending problem than have to respond to a critical equipment failure after it happens.

Which is why many facilities are expanding their thermal imaging practice to include "thermal monitoring" with semi-fixed, always-on wireless infrared sensors.

In the past, troubleshooting intermittent problems required a technician to be in the right place at the right time, which was not always possible. Installing semi-fixed thermal imaging sensors on multiple components provides technicians with a more comprehensive thermal view of equipment assets simultaneously, in real time, over long periods of time, making them much more likely to catch intermittent faults in the act.

Thermal imaging sensors simultaneously capture thermal images of multiple motor, fan, pump and conveyor components to help spot abnormal thermal patterns in bearings, shafts, casings, belts, gearboxes and other parts.

Technicians can then compare these images across an entire duty cycle to see what else is going on when warning signals appear. They can then respond proactively to problems and head off potential failures in other equipment down the line.

The thermal monitoring sensors provide more data to help maintenance personnel better determine whether they need to address the problem immediately or whether they can wait for off-peak hours or until the next regularly scheduled maintenance period.

This flexibility in determining the issues before an accident saves time, saves money and saves lives.

In addition to preventing equipment failure, predictive maintenance also leads to more efficient maintenance. First, this strategy makes it easier to schedule maintenance when it is necessary, before problems build up. This could save tens of thousands of dollars in costs.

Predictive maintenance also eliminates unnecessary inspections, which translates to fewer wasted resources.

The ability to schedule maintenance for the most convenient time, before emergency maintenance or repair becomes necessary, allows for less planned downtime of a piece of equipment vital to a mine. It also makes it easier to plan ahead. For example, a foreman can order spare parts before they are needed.

Maintenance teams can perform multiple procedures while the line is down for repair, to maximize efficiency.

The benefits to having an integrated system that can monitor and detail the status of equipment is only further improved with AI.

Machine learning takes condition monitoring data and automatically defines a machine's baseline conditions and sets thresholds for acute and chronic conditions so that you know in advance--and with confidence--when your machine will require maintenance.

After installing a vibration sensor into equipment, the collection of enough data to establish a baseline for the machine is required. Machine learning removes the chances of human error by automating the data analysis.

A condition monitoring solution with machine learning will recognize the machine's unique baseline of vibration and temperature levels and automatically set warning and alert thresholds at the appropriate points. This makes the condition monitoring system more reliable and less dependent on error-prone manual calculations.

When a vibration or temperature threshold has been exceeded, a smart condition monitoring system provides both local indication, such as sending a signal to a tower or remote alerts like emails or text messages. This ensures that warnings are addressed quickly.

In addition, a condition monitoring solution that allows you to log the collected data over time enables even more optimization. With a wireless system, vibration and temperature data can be sent to a wireless controller or programmable logic controller for in-depth, long-term analysis.

As the next logical step towards an integrated smart mine, the condition monitoring sensors, when paired with an AI, is even further automated and efficient by being fully adapted via the internet of things (IoT).

Mesh networking

With a current IoT system, utilizing established Wi-Fi and cellular connectivity, automated equipment can function independently from human interaction as the devices themselves are "aware" of the functions of the others.

Sweden-based SKF has taken it a step further through its development of advanced sensors that allow businesses to incorporate condition monitoring and creating an uninterruptible network for stability and security of the communication signal.

With the intercommunication of sensors, AI algorithms, programmed pathing and any other data being transmitted, an incredible amount of information is whizzing between these heavy-duty computers on wheels.

To better ensure the security and stability of the data, a system called mesh networking is being realized.

Simply put, mesh networking is the output of wireless signals being transmitted from connected devices. This would be like the router in your home, which transmits the signal for internet, connecting to your phone, which does the same thing as the router, which is also connected to your television and so on.

This method ensures that in the event of malfunction or loss of signal, communication between all devices persists.

This means that data can be routed around radio obstacles, such as pipework and liquid storage vessels, that create signal blocks for conventional line-of-sight systems and sent over greater distances than would be possible using a single device.

"The mesh network is self-forming, which makes it easier and quicker to deploy than other wireless communications technologies such as Wi-Fi or Bluetooth," SKF Product Line Manager Chris James explains. "It is also innovative in the way it manages available bandwidth and the power consumption of the sensors, which leads to a long enough battery life to meet the needs of our multi-year service contracts."

"Critically, the sensor gathers data consistent with our manual data-collector, particularly when it comes to detecting early-stage bearing defects. Although severely damaged bearings are relatively straightforward to detect, by that stage they are close to failure – the key is to find defects early, so that corrective action can be planned in good time with minimal disruption," he added.

SKF has released a compact and cost-effective vibration and temperature sensor for monitoring the condition of rotating parts on heavy industrial machinery.

Designed principally for use as part of an SKF Rotating Equipment Performance (REP) solution, the sensor - called the SKF Enlight Collect IMx-1 - enables customers to reduce both expensive unplanned downtime and their maintenance costs.

The autonomy of heavy equipment in mining and other industries is around the corner and a method of error detection is part and parcel of a stable, efficient and safe operation.

And with newly developing autonomous mines being built around the globe, establishing a system for low downtime and high productivity leads them to seek similar technologies for condition monitoring.

As technology for interconnected devices becomes more advanced, the efficiency and cost and benefits to utilizing that technology becomes more obvious. It may be that in the next decade we'll have fully automated mines with a single "George Jetson" punching away on a button to spit out sprockets.