What is unsupervised learning? Unsupervised Learning (see wikipedia) is used when no historical data is available that involves past faulty behaviour and/or ageing progression. Therefore,
Railway point systems
Railways point systems’ command signal delivers critical information on system’s health and reveals defects and aging phenomena.
Electric current from the railway point system activator is captured. A dataset is available describing healthy and faulty behaviours. Need is to build a model that automatically detects faulty behaviour to raise an alert to maintenance operators.
DiagFit was used to build a predictive model to alert on faulty behavior. Both supervised and unsupervised (without faulty data) learning were possible.