Adaptive Anomaly Detector

THE POWER BEHIND YOUR CONDITION MONITORING & PREDICTIVE MAINTENANCE SOLUTION

Optimum Detection Capability with Low False Alert Rates

Humaware’s Adaptive Anomaly Detector provides users with trusted actionable information to optimise maintenance schedules and increase asset availability.

The Adaptive Anomaly Detector identifies deviations as anomalies; these include events, discontinuities and trends in the data. Early detection of these types of anomalies is essential for providing enough time to plan and perform maintenance operations during the estimated remaining useful life of the asset.

it is an essential building block for implementing predictive maintenance.

The Adaptive Anomaly Detector offers:

  •  Low false alert rates
  • Optimum detection capability
  • No setting & Maintaining individual fixed thresholds
  • Reliable data for remaining useful life estimates
  • Not machine or asset specific
  • Use any time series data
 

A Solution You Can Depend Upon

Humaware’s Adaptive Anomaly Detector allows you to create competitive, effective and scalable Condition Monitoring and Predictive Maintenance solutions.

Standalone Solution

The Adaptive Anomaly Detector is available as a stand-alone solution. APIs allow you to interface with the cloud-based Adaptive Anomaly Detector and view the results in your own interface.

Extensively
Proven

Humaware’s technology has been extensively proven in the rail and aerospace sectors and can be applied to any application which requires the analysis of time series data.

Meeting Predictive Maintenance Needs

Accurate and early detections are are essential for the implementation of predictive/prescriptive maintenance and the generation of remaining useful life estimates.

End-to-End
Solution

SmartVision™ by EKE-Electronics is powered by Humaware’s Adaptive Anomaly Detector to provide a multi-fleet remote diagnostics & predictive condition-based maintenance system for train and track.

Solve Your Threshold Management Problems

The application of Condition Monitoring/Predictive Maintenance to a large population (or fleets) of assets often results in reduced effectiveness of your current solution. Typical issues include:

HIGH NUMBER OF MISSED ALERTS

DATA VARIATION ACROSS ASSETS

DEPENDENCIES ON OPERATING CONDITIONS

DATA VARIABILITY/STABILITY

HIGH FALSE ALARM RATE

HIGH DATA MANAGEMENT

Many feel that they are left with no choice but to apply individually tailored fixed thresholds for each individual asset in a population. This may not be a problem for you right now, but as your solution scales, the task of setting and managing individual fixed thresholds across hundreds of assets can quickly become unmanageable, distracting valuable resource and extending the time to detect failures.

Humaware’s Adaptive Anomaly Detector provides the solution to automate key stages of the process to achieve optimal system performance