Skaiwatch Enables Smart Maintenance
Skaiwatch Harnesses The Power of Machine Learning
The potential of cognitive predictive maintenance which can also be called smart maintenance,
can only be fully harness by advanced statistical methods and machine learning. Skaiwatch uses
the state of the art algorithms available for industry 4.0, to detect early sign of failure,
anomalous state of machines, learn the anomalous behaviour, and alert the user when such behaviour are exhibited by machines.
Skaiwatch combines the traditional vibration analysis with deep learning to detect early warning and pin point to the related asset.
The key technologies behind the workflows
Secure transfer of data from different sources via APIs
from partners or via secure file transfers.
Certified data quality analysis service via DNV GL’s Veracity Platform is available.
Signal Processing Libraries: Transforms signals to increase efficiency of analysis.
Early Warning and Prediction: Alerts when predictive signals are found or anomalies
are detected using machine learning, deep learning and transfer learning algorithms.
Root Cause Analysis: Checks the entire dataset to find best possible cause which explains the issues and anomalies.
Remaining Useful Life Calculation. Statistical methods find the number of hours/days of normal operation remaining for an asset.
Mobile and customized visualization using PowerBI and javascrip to show alerts and Key Performance Indicators.