Predictive Maintenance
Production that “thinks along” with you
Receive a notification that a machine cylinder needs to be replaced shortly before it fails. And be assigned a time slot by the ZELTWANGER maintenance service for them to come and replace the defective part.
This is forward-looking production – predictive maintenance from ZELTWANGER.
Increased revenue and reliability – predictive maintenance has quantifiable advantages. Those who can foresee when parts in systems and machines need to be exchanged before they break down avoid downtimes and situations where defective parts lead to dangerous misinterpretations.
By evaluating the data of all sensors and actuators in our systems, many problems can be detected before they actually happen, e.g., temperature fluctuations in the power supply, changes to compressed air values, or the number of switching cycles.
It is possible to mathematically calculate when the service really starts to pay off. This means that investment costs can be reduced thanks to targeted maintenance because parts are only replaced when it is really necessary.
Energy costs are also reduced because predictive maintenance controls the switch-off of the supply air during (process) breaks.
At the same time, machines that are not at a standstill have a higher output and plant availability.
Predictive maintenance is the cooperation of system hardware and software. This allows processes to be analyzed and data to be saved and evaluated in a Cloud.
The results of condition monitoring are displayed on a dashboard and you receive a focused summary of thousands of measured values.
Even the smallest changes to processes are shown and the timing for system maintenance can be predicted.
All process that take place in systems can be monitored and assessed. Some of these include:
Lots of data is collected for predictive maintenance. It is essential that this data is stored securely. The cloud server of our partner Festo is based in Germany and is accessed via secure web gateways.
Artificial intelligence in machines mainly involves machine learning. Systems evaluate the data collected and over time they are able to adjust settings automatically when conditions change.
This is particularly interesting when it comes to predictive maintenance because the machines learn to automatically adjust predictions on the basis of data evaluation when circumstances change.
This applies, for example, to temperature differences, air humidity, or stronger or weaker vibrations (if, for example, new machines are set up in the area). The system can also incorporate material changes and optimize cycle times.