Open AccessStatistical Anomaly Detection for Train Fleets
Author(s) -
Holst Anders,
Bohlin Markus,
Ekman Jan,
Sellin Ola,
Lindström Björn,
Larsen Stefan
Publication year - 2013
Publication title -
ai magazine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.597
H-Index - 79
eISSN - 2371-9621
pISSN - 0738-4602
DOI - 10.1609/aimag.v34i1.2435
Subject(s) - anomaly detection , train , anomaly (physics) , computer science , bayesian probability , data mining , parametric statistics , event (particle physics) , artificial intelligence , statistics , mathematics , geography , physics , cartography , condensed matter physics , quantum mechanics
We have developed a method for statistical anomaly detection that has been deployed in a tool for condition monitoring of train fleets. The tool is currently used by several railway operators across the world to inspect and visualize the occurrence of “event messages” generated on the trains. The anomaly detection component helps the operators quickly to find significant deviations from normal behavior and to detect early indications for possible problems. The method used is based on Bayesian principal anomaly, which is a framework for parametric anomaly detection using Bayesian statistics. The savings in maintenance costs of using the tool comes mainly from avoiding costly breakdowns and have been estimated to be several million Euros per year for the tool. In the long run, it is expected that maintenance costs can be reduced by between 5 and 10 percent with the help of the tool.