3 min read

  • Vágólapra másolva

Case study: Anomaly detection, predictive maintenance

The state-of-the-art solution for anomaly detection rests on an AI system solution based on unsupervised learning that, by learning past usage patterns and general operational routines, continuously monitors unexpected, out-of-the-ordinary events in the system by forming user or operational groups, so-called clusters, from these patterns. It then determines the severity grade of the event compared to predefined standard levels and sends out a corresponding alert. In the following article, we analyse an applied case of this technology:

Futuristic,Concept:,Blurry,Factory,Digitalization,With,Information,Showing,Efficiency,Percentage

Predictive maintenance can be used to optimise the operation and capacity utilisation of production equipment by analysing the sensor data of individual parts or components. By combining these data with production plans, an artificial intelligence-based anomaly detection system can predict the lifetime of each component or anticipate failures, thus ensuring continuous operation and optimising maintenance procedures.

Stratis Esettanulmany Karbantartasi Strategiak 768x431

Source: https://www.ixon.cloud/applications/predictive-maintenance.

ases the efficiency of the production plant, as there no longer occurs any unexpected downtime in the production chain, while costs are reduced, as the replacement of individual parts happens only when it is really necessary. To implement predictive maintenance, a manufacturing plant must have suitable infrastructure, but building this infrastructure will at the same time offer additional benefits to operate a "smart manufacturing plant".

Stratis Tablazat 768x240

If the above conditions are met, the manufacturing plant can use artificial intelligence to optimise its processes. Predictive maintenance and related technologies are therefore essential for the creation of a "smart manufacturing plant" that can operate with both higher efficiency and lower costs than its conventional counterparts.

By using predictive maintenance alone, a manufacturing plant can reduce its maintenance time by 20-50%, while the uptime of manufacturing equipment maintained in this way can be increased by 10-20% and the overall maintenance costs can be reduced by up to 10%.

Meanwhile, you can increase your productivity by up to 15% thanks to increased efficiency and the elimination of downtime caused by defects and maintenance.

Source: https://hbr.org/2016/05/where-predictive-analytics-is-having-the-biggest-impact

What business problem
can we help you solve?

Left hand art Right hand art

You may also like these