FUTURE MANUFACTURING
Sources : elunic AG
Automated anomaly detection reduces maintenance costs . result , the time taken to detect a significant anomaly increased to 28 days before an anticipated failure , giving the operator ample warning time to perform a maintenance task .
Increased plant effectiveness , lower maintenance costs
Specialist personnel receive error messages directly on their smartphones .
AI also detects complex errors in rule-based systems
With rule-based systems , users would have to determine in advance for numerous combinations of deviations which threshold values are still permissible in any given case . Due to the sheer amount of conceivable combinations , this is impractical to manage or , at best , only possible in the case of relatively simple manufacturing processes . AI-based anomaly detection , therefore , offers higher detection reliability compared to rule-based systems while requiring less development effort .
Warning many weeks before failures
The use of anomaly detection at Certuss Dampfautomaten GmbH & Co . KG provides an illustrative example : During the project , a pressurized steam boiler monitored with an anomaly detection solution from elunic AG was reporting a combination of pressure and temperature shifts . As a result , the system predicted that the boiler would likely fail within 48 hours . With the support of elunic ’ s data scientists , this was followed by more intensive training of the AI model to further optimize early detection performances . As a
This example demonstrates the benefits of implementing anomaly detection and improving overall plant effectiveness . The performance of plants enhanced by anomaly detection increases while , at the same time , maintenance costs decrease . This is achieved through reduced downtime and fewer production errors , all the way to zero-defect output .
Often , anomaly detection systems for manufacturing companies are developed as a component of existing IIoT software modules with the option to connect them to ERP , BDE , and MES systems via various interfaces . In this way , all production-related areas of the companies receive information about the condition of the plant and equipment . l
Claudio Gusmini Director IoT Solutions elunic AG