IoT and AI-based Solutions for detecting issues early before they affect your business 

From unexpected failures to planned repairs  

Our customers are taking the next step in industrial operations and maintenance with data, analytics and machine learning. Combining Anomaly Detection from existing process data with Smart Sensors and Industrial Analytics (MP INtelligence & IoT Flex) – our customers achieve less unplanned downtime, more stable processes and highest output quality.

Moving into predictive and prescriptive maintenance means O&M teams require new insight into the production processes and asset condition. Our unique approach combines machine learning analytics, advanced IoT sensors and our service team of data scientists, mathematicians and process industry experts.

Combination of technological capabilities with industrial process experience differentiates us from pure technology vendors. We run operation and maintenance activities for our clients. We know what works in real life and what doesn’t. Our offerings have been tried and proven in actual use.

IoTFlex – Smart Sensors for measuring actual condition 24/7  

Caverion’s IoTFlex uses smart sensors to measure the actual condition of machines and employs real-time analytics to predict the occurrence of possible faults.

IoTFlex enables easy adjustment of the failure prediction based on the running method of the machine. The system adapts easily to a range of measurement needs, such as air particle measurements, vibrations, trace heating, and condition measurements and oil analyses for electric motors. We have also installed IoTFlex applications with wireless sensors as well as video and audio analytics.

The IoTFlex service concept contains:

  • Smart sensors
  • A remote management device with mobile interface
  • Analytics through a measurement and analysis application developed by Caverion
  • The special expertise of Caverion’s experts

MP INtelligence - Anomaly Detection with existing process data

MP INtelligence identifies process issues early from existing process data with advanced machine learning algorithms and high-performance computing. This means actions can be taken early, resulting in less unplanned downtime, improved output quality and greater process stability. 

Machine Learning algorithms use the plants existing process data and are designed to handle thousands of measurement points with high-frequency data resolution. The service combines unsupervised (automatic) and supervised (user feedback) machine learning to ensure the mathematical models are continuously updated. 

MP INtelligence is actively used by several power plants, factories and production facilities across the Nordics since 2016. 

The most important qualities are

  • real-time collection and analysis of thousands of measurement points
  • early detection of equipment failures and process disturbances
  • detection of new types of anomalies as they occur for the first time
  • Web-based UI for handling alerts
  • Unique support model with Caverion experts  


  • Typically reduces the costs of unexpected downtime by 30 - 50% 
  • Streamlines maintenance with more time to diagnose and plan corrective actions
  • Maximizes output quality and asset lifecycle
  • Fast deployment and time-to-value
  • Secure, scalable and proven solution