Office for Technology Commercialization
http://www.research.umn.edu/techcomm
612-624-0550

Anomaly Detection Algorithm Uses Hybrid-Logic to Predict System Failure

Technology #20110023

Questions about this technology? Ask a Technology Manager

Download Printable PDF

Collections
Image Gallery
Anomaly DetectionFault DetectionHybrid-Logic Algorithm
Categories
Researchers
Jaideep Srivastava, PhD
Professor, Department of Computer Science and Engineering, College of Science and Engineering
External Link (dmr.cs.umn.edu)
Managed By
Andrew Morrow
Technology Licensing Officer

Anomaly Detection Predicts System Failure

The anomaly detection algorithm assesses the overall probability of systematic failure in complex systems. The fault detection algorithm uses both physical and statistical data in tandem to assess the systematic risk value which reflects the overall probability of system failure in complex systems. The overall risk measure is a disjunction over all failure modes of the system.

MN-IP Try and Buy
Try
  • Trial period of 6 to 12 months. $5000/6 months.
  • Fee waived for MN-based companies or if sponsoring $50,000+ in research.
Buy
  • Exclusive license for a $30,000 conversion payment.
  • No patent costs.
  • Royalty rate of 3% (2% for MN company) after first $1 million in product sales.

** View the Term Sheet **
** Contact Andrew Morrow for more information.

Hybrid-Logic Algorithm Combining Statistical and Physical Data

Unlike current systems which use statistical inputs to derive failure probability, the hybrid-logic based algorithm combines statistical data with domain information, causal and functional, to diagnose complex anomalies. The described algorithm not only works well for simple and causally complex systems, but also for functionally complex systems where current state of the art Bayesian networks often fail to detect potential anomalies. Potential applications include vehicle health management (e.g avionics) and device maintenance (e.g biomedical).

BENEFITS OF HYBRID-LOGIC ALGORITHM FOR FAULT DETECTION:

  • The algorithm for fault detection uses both physical and statistical data to assess the probability of failure in complex systems.
  • High performance levels in simple, causally complex and functionally complex systems.
  • Potential applications include vehicle health management (e.g avionics) and device maintenance (e.g biomedical).