Predictive Maintenance: Technologies and Advatages

In an industrial plant, decreasing the probability of failure of any component of the production chain as much as possible is of considerable importance. The malfunction of a single component, can lead to the stoppage of the production process, an event that can have a strong impact on the business of the activity.

The classic methods of preventive maintenance are designed to protect the system from these potential forced stops through scheduled overhaul and maintenance of components at regular time intervals. This methodology, while effectively reducing the probability of failure of the machines, makes a recording at long intervals of time. With Industry 4.0 and the increasingly advanced digitalization of processes, it is possible to have an enormous amount of data, resulting from constant monitoring, regarding the performance and status of industrial components. Starting from this amount of information, it is possible to put into practice a predictive maintenance approach, based on the real operating conditions of the assets.

Advantages of predictive maintenance

With predictive maintenance, interventions are planned and based on the data collected by monitoring; specific learning models exploit this data to estimate the probability of machine failure. The result is an almost continuous resolution of component monitoring, which leads to a decidedly more effective planning of maintenance interventions in economic terms.

In this way, the risk of breakage, with consequent replacement, of a component is avoided as much as possible and the interruption of processes can take place in an even more functional way by planning maintenance interventions in the best possible way.

Furthermore, the life cycles of the individual assets are considerably increased and a considerable improvement is also achieved in terms of safety at work.

Types of predictive maintenance

Depending on the physical principle on which the monitoring of system components is based, there are different predictive maintenance technologies. Here are some examples:

• Vibration analysis: mainly applied for monitoring rotating machines (motors, pumps, gearboxes) which in their operating condition generate vibrations at certain frequencies. If the frequencies detected, using specific instrumentation (permanently installed on the machine or portable, for monitoring at regular time intervals), are outside the nominal operating range, there is an anomaly in the machine, from which it is possible to predict when the probable failure will occur.
• Ultrasound measurements: predictive maintenance based on ultrasound detection is designed to analyze and identify anomalies that emit sounds that are not audible to the human ear. In this case, the instrumentation used detects sound signals with frequencies above 20 kHz generated, for example, by anomalies such as air leaks (anomalies on pressure circuits) or problems on hydraulic valves.
• Industrial Thermography: predictive maintenance technology aimed at detecting temperatures and thermal imbalances. By using a thermal imaging camera, it is possible to detect the operating temperatures of a specific component and, consequently, to compare the temperatures detected with the nominal operating temperatures. In case of detection of abnormal overheating, possible causes of irreversible damage to the machines, it is therefore necessary to plan a maintenance intervention.
• Endoscopic video inspections: they allow you to carry out visual analyzes in points of the machine components that are not easily reachable by the human eye. Through high resolution cameras, in fact, it is possible to reach difficult to inspect points inside components such as turbines, exchangers, bearings, compressors and more. The high resolutions are able to capture images and make videos very accurately, from which it is possible to identify any damage or anomalies compromising the operation of the machine.

All predictive maintenance technologies listed are intended to make routine maintenance more efficient and to have each component of the system in the best working conditions.

ISE has a multi-purpose expertise with respect to predictive technologies that allows to identify and analyze potential faults from multiple points of view, thus increasing the probability of detection of the same fault even in its initial manifestations (Early Fault Detection).