The Predictive Intelligence software module, part of the PROGNOST®-NT (version 21) monitoring and diagnostic system, is designed to forecast the remaining useful life of valves and packings. By leveraging AI-driven predictive algorithms and real-time data, we boost equipment availability, reduce unplanned downtime, extend maintenance planning lead times, and ensure full utilization of component lifetimes.

Included in our service level agreements, we offer expert support based on data-driven insights.

Valves and packings are key components that often cause unscheduled compressor downtimes. Predicting their lifespan and compressor availability gives more time for maintenance planning, allowing these costly parts to be used fully.

• Optimized Maintenance Planning: Coordinates maintenance with other scheduled periods, minimizing disruptions.
• Maximized Component Life: Ensures components last longer without compromising reliability.
• Extended Availability and Reduced Costs: Enhances maintenance planning for extended availability and minimized costs.

• AI-Based Predictions: Predictive Intelligence uses AI and real-time data to predict the remaining useful lifetime of key components.
• Enhanced Productivity: Combines real-time data with historical information to establish the Projected Availability Span and Remaining Useful Lifetime.
• Focused Maintenance: Assesses the expected lifetime of valves and packings to improve compressor availability.
• User-Friendly Reports: Provides easy-to-understand reports indicating the best time for maintenance.
• Secure Data Analysis: Continuously analyzes data in a secure Azure cloud environment.

The module uses two independent algorithms, each utilizing different signal sets and calculation methods. This independence enhances the precision and reliability of the results.

Remaining Useful Life (RUL)

The Remaining Useful Life feature predicts the lifespan of packings and valves at the cylinder level using vibration signals. It applies to both horizontal and labyrinth compressors, providing a maximum forecast of 58 days. Initial results are available after the first anomaly is detected in the data set.
This feature is patented under EP2023/065418.

Projected Availability Span (PAS)

The Projected Availability Span is a feature that predicts the operational reliability of sealing elements in compressors. It uses pressure and temperature signals to calculate the timeframe during which these components are expected to operate without anomalies. This applies to SV/Packing or DV for all cylinders, including labyrinth compressors. Initial results are available after 101 days of continuous operation.
This feature is patented under EP4369127A1.