Condition monitoring as the basis for smarter maintenance planning

Industrialisation began with the invention of the steam engine. So many years later, we are in the midst of the fourth industrial revolution. Machines are now intelligent and connected to each other. Data is captured, processed and evaluated in real-time through artificial intelligence. 

A central task in an Industry 4.0 environment is to further drive up machine availability and output via digital solutions. Collecting, storing and evaluating data from different machines and systems, as well as building digital twins, are the prerequisites for comprehensive maintenance concepts. Data are the raw material of the future. An optimised maintenance strategy based on the health of your plants requires permanent monitoring, also known as condition monitoring. The intelligent and communicative solution for security and energy distribution REX system helps collect data very specifically over the entire 24 V range, the heart of any automation solution.

What maintenance strategies exist?

We can broadly divide maintenance into two approaches: reactive maintenance (when there is a failure) and planned maintenance. At first glance, plan A seems the easiest and cheapest. But those who consider the full picture will see that unplanned downtime quickly leads to greater expenditure. In comparison, machines are down for longer, it will take longer to find the cause, and the current bottleneck in the spare parts market also plays its part. So better schedule maintenance in advance, and there too you have several options. Preventive maintenance relies on empirical values and predetermined intervals. A condition-based maintenance methodology works with measurements that determine and monitor the health of the machine. If the measured values deviate, an alarm follows. Finally, predictive maintenance combines both approaches. Based on a variety of data and empirical values, artificial intelligence predicts which components will fail and when. This then results in a very precisely devised plan
of maintenance tasks. 

Condition monitoring and its benefits

In the industrial arena, machinery maintenance is often done preventively to minimise downtime. Parts prone to wear are replaced at predetermined intervals. However, this is not the most efficient use of components, making it a relatively expensive operation. Operators therefore increasingly prefer other concepts to control their maintenance costs. Condition monitoring technology continuously monitors the state of the machine and delivers the data relevant for maintenance planning by identifying the actual state of wear. Before they reach end-of-life, components can thus be replaced while reducing maintenance costs. Intelligent condition monitoring adds a number of benefits. Improved process stability leads to higher machine availability, and thus to more produced quantities. Condition monitoring detects anomalies and wear at an early enough stage for sustainable optimisation of all maintenance tasks.

Digital protection with condition monitoring

For a digital connection to the upstream control system, machine operators rely on fieldbus solutions. Think Profinet, EtherCAT, Ethernet-IP, Modbus-TCP, but also the fieldbus-independent communication of IO-Link is increasingly common. The REX system provides the basis for smart maintenance planning. The 24 V condition monitoring solution combines selective overload protection, compact energy distribution of load circuits, and transmission of measurement data via the Control®Plex controller. These controllers are able to apply all conventional communication protocols and transmit all kinds of diagnostic information for the entire 24 V range to the upper level. This includes information on input voltage, load current, load voltage and threshold values, as well as various settings, such as workload warning thresholds, in percent. The measured values and continuous transparency lay the foundation for a condition-based maintenance plan. The measured values measured during system start-up form the basis for target and warning thresholds. The actual values to be derived from these are compared with the basic measured values. If there is a deviation, an error message is displayed. Then add an intelligent algorithm (artificial intelligence) and the operator's knowledge and experience, and you have everything you need to take the step towards predictive maintenance.■