What is Predictive Maintenance: Definition, Applications and Examples

Predicting unplanned machine failures and identifying maintenance needs in time – predictive maintenance is revolutionizing the industry and is a crucial step toward a proactive and intelligent maintenance strategy.

But what exactly does the term mean? Predictive maintenance involves monitoring equipment for early signs of irregularities before they lead to failures. Unlike fixed maintenance schedules, predictive maintenance intelligently forecasts needs. Indicators of this intelligence can include increased operating temperatures or higher energy consumption.

In the context of Predictive Maintenance, the term "preventive maintenance" is also frequently mentioned. These maintenance strategies are particularly relevant for manufacturing companies, as their value creation depends on the smooth operation of equipment and machinery. One can hardly imagine the cost of a one-minute production halt for one of the major German automobile manufacturers. Therefore, the question of whether investing in intelligence for preventive measures is worthwhile is quickly answered.

Predictive Maintenance offers companies the opportunity to significantly enhance their operational efficiency through the following advantages:

• No unexpected production downtime: This is perhaps the greatest value that Predictive Maintenance provides, as unexpected downtimes can result in substantial financial losses. However, it is not the only benefit, which makes predictive maintenance so attractive.
• Efficient workforce planning: In times of skilled labor shortages, professionals capable of maintaining specific machinery are becoming increasingly scarce and expensive. With Predictive Maintenance, for example, when an increased operating temperature indicates a specific maintenance need, specialists no longer need to visit the site on suspicion but only when there is actual work to be done.
• Reduced material costs: The consumption of wear parts and consumables can also be managed more sustainably. For instance, if a saw needs a new blade, increased coolant consumption will indicate this need. Thanks to Predictive Maintenance, the saw blade can be replaced at the exact right time, preventing unnecessary waste of coolant. 

Now, what do you need for Predictive Maintenance? The first requirement is sensors that measure various parameters. These sensors may already be integrated into the device or can be retrofitted. The second step involves transmitting the measured values to an online platform. Depending on the application, a cost-effective Wi-Fi connection, a reliable 4G/5G connection, or a long-range LPWA connection might be most suitable. In such an IoT network, many parameters like power supply, network coverage, and security requirements need to be considered.

Once the data arrives on the online platform, the magic happens. For predictive maintenance, threshold values can be defined, the exceedance of which triggers a maintenance appointment with specific tasks. This allows maintenance activities to be managed as needed, aiming to minimize time, resources, and downtime.

Similar to Ziehl-Abegg, RUD Ketten Rieger & Dietz GmbH & Co. KG is also a manufacturing company.
The leading manufacturer of high-performance chains and industrial conveyor technology relies on Predictive Maintenance through the Internet of Things (IoT). Telekom's IoT cloud enables RUD to remotely monitor conveyor chains in power plants and industrial facilities. Sensors on the equipment send data on load, wear, and heat generation to the cloud, where it is analyzed and visualized.

Thanks to Predictive Maintenance, RUD's customers can monitor the condition of the chains in real time and plan maintenance precisely. This is particularly helpful in the harsh industrial environments where RUD chains are used. Excessive heat, dust-covered rollers, or unusual speeds can cause the chains to malfunction. Predictive Maintenance minimizes unplanned downtime and significantly reduces maintenance costs, thereby increasing the efficiency and reliability of the equipment.

So, everything perfect with Predictive Maintenance?As with many digitalization projects, one might affirm this question, if it weren't for the implementation costs. Of course, the installation of sensors and setting up the dashboard is labor-intensive. However, there are already initial approaches to optimization in this area. The solution is called Dormant Connectivity, which is a cost-effective way to network devices. With Dormant Connectivity, it is worthwhile to monitor not only large production lines in industrial companies but also smaller devices such as cleaning robots.