Modern AI and FMEA
APM 360™ is focused on optimizing the performance of assets to increase reliability and availability, minimize costs, and reduce operational risks. Outcomes include increased asset availability by as much as 5%, reduced reactive maintenance by 10-40% and up to 10% in inventory cost reduction. APM 360™ leverages machine-learning based on artificial intelligence to detect anomalies and take into account complex, dynamic behavioral machinery patterns and contextual data relating to the manufacturing process at large.
Symphony Industrial AI’s APM 360™ was selected the GOLD WINNER in The 19th American Business Awards®. APM 360™ was voted Gold in the New Product IoT Analytics Solution Category. APM 360™ and Performance 360™ has been voted the best new products of 2020 by the readers of Oil & Gas Engineering.
Click here to read more >>
Key Technology Differentiators
- Harnessing of Process and high frequency vibrations to AI models for high fidelity predictions and wide fault coverage
- Unsupervised AI models to detect unknown unknowns and not just known defects based on pattern recognition
- Self-tuning adaptive AI models that self-train based on auto-detection of maintenance events and/or major process changes
- FMEA Engine for cause advisory and actionable recommendations
APM 360™ Key Features
Asset Health Intelligence
- Asset Health Score for asset and plant
- Asset Performance Degradation
Real-time Asset Monitoring
- Asset templates for rotating & static equipment
- Predictive Analysis, early detection of asset anomaly
- FMEA Library for cause determination and corrective action
- Predictive maintenance advisory
- Maintenance KPI prediction, i.e.: MTBF, MTBM
- Digital twins for What-if analysis – run-time extension, failure prediction
SAAI’s Asset Templates cover major assets:
- Heat Exchangers
APM 360™ Architecture
Case Example One: APM for large multi-stage centrifugal compressors
Maximize availability of integrally geared multi-stage centrifugal air compressor in an 3000 tons / hour Ammonia plant by predicting anomalies with cause and mitigation action advisory. Digital twin used blend of unsupervised with supervised ML models – TDA, LSTM encoders, KNN, Decision trees for anomaly detection. Benefit delivered $2M in a year by preventing unplanned outage events.
Case Example Two: Predictive Maintenance of LNG compression train
Minimize unplanned downtime of multi-stage centrifugal fuel gas compressor, gas turbine driven export gas centrifugal compressor, reciprocating boil-off gas compressor trains in a large LNG liquefaction facility by detecting incipient equipment faults & their causes. FMEA-powered blend of unsupervised & semi-supervised ML: TDA, SVM, NN Autoencoder, Random Forest. Benefit delivered over $2M through timely repair recommendations and outage avoidance.
We share insights , case studies and analytics to help you reduce unnecessary costs and focus resources on critical assets. Our data science team can also help find these hidden opportunities within your historian time-series data, EAM or CMMS data you’re already collecting.Schedule APM Webinar
Built on the Eureka AI Industrial Platform
Eureka AI Industrial Platform is the foundation layer of APM 360™. Beyond secure data connectivity, processing and storage, Eureka Industrial Platform provides rich user features such as connectors, analysis, dashboarding, alerts, cases and workflows that are seamlessly integrated into the APM user experience.