Convergence of AI/ML and Digital Twin Technologies in Predictive Maintenance
As our customers continue their digital transformation, we start to see more EAM customers adopting AI and digital twin technologies together in their predictive maintenance (PdM) programs. One of the best examples is Engie, a global leader in low-carbon energy and services who won a SAP Innovation Award 2022. Engie realized impressive benefits: 11% reduction in product loss and 35% maintenance efficiency gained for those photovoltaic equipment monitored and analyzed by SAP Predictive Asset Insight (PAI) and SAP Enterprise Productive Development (EPD).
The convergence of AI/ML and digital twin-based simulation makes a very compelling PdM use case. Based on the sensor data ingested from devices or data layers, the digital twin technologies simulate “virtual sensors” that would be otherwise physically/economically impossible to instrument. A complete picture of the equipment is formed when the virtual sensors and physical sensors readings are blended and analyzed together by the AI. Maintenance rules can then be set up by business users to continuously monitor the equipment and generate alerts and/or EAM notifications when an anomaly is detected, or a failure is predicted by the AI.
While Engie has implemented this use case using PAI and EPD, this use case is now supported by SAP Asset Performance Management (SAP APM, the successor to SAP PAI) with standard integration to EPD. Moreover, APM is slated to have a much-improved anomaly detection capability in Q3. This new and enhanced anomaly detection capability is based on SAP AI Core and provides a no-code user experience and explainability for reliability engineers. I will share more details around APM’s the new AI capabilities in future blogs so please stay tuned.