Applications
From Physical measurement to Asset Diagnostic
Our complete asset diagnostic solution uses an AI module - based on a physical neural network - to analyse the signals and display the ”machine condition” via an intuitive dashboard.
The raw data are collected by analog sensors placed at critical points on the equipment and transmitted through our proprietary data acquisition system.
Our technology
Measurement of raw signals
Propriétaire AI System
Our Data Management
Accuracy & Performance
Other methods for anomaly detection, based on
standard machine learning, commonly use partial
data (less than 100Hz) or global values (e.g. RMS).
By not taking advantage of the full raw signals
analysis, they’re limited to simple problems such as
bearings / simple rotating machines
VS
Our system provides early alerts and a complete
understanding on the physical phenomenon involved, by comparing the entire analog signals to a period of time assumed to be representative of normal conditions (adjusted over time).
By exploiting the full raw signals and using a wide range of sensors, we offer a universal solution that can address any type of equipment across any industrial sector.
Present approach
-
Small sensor density
-
Scarce data – not real time
-
Large amount of data to transfer
-
Human intensive task
Our solution
-
High sensor density
-
The complete data – real time
-
No data transfer
-
No human intervention
MAG.NET
Human hearing for predictive maintenance
