Step 1: Digitizing & Filtering
Preprocessed Signal up to 10 MHz sample rate
Raw signal for post processing and/or history database
Step 2: Feature Extraction
Several parallel Real Time processing paths in Time and/or Frequency domain
Key Performance Indicators to DCS, Process Information & Control Systems and controll room displays
Step 3: Machine Learning / Artificial Intelligence
Customized Real Time Processing, made in Python and C++
ML and AI algorythms Results compining AE and Process parameters
Real Time 24/7 Predictive Condition Diagnostics.
- Rotating / moving machine elements: Friction level, stress level, changes & stability, Cracking.
- Pressure vessels, valves & pipes: Cracking, leaks, material stress level.
- Pneumatic and Hydraulic systems: Cracking, leaks.
Real Time 24/7 Predictive Process Diagnostics.
- Flow control: changes & stability
- Fluidized Beds: Process changes & stability
- Process layers: Level control
- Chemical processes diagnostics