As a preliminary stage to "predictive maintenance", anomaly detection is used, which uses machine learning to recognize patterns in large data sets. It is useful, for example, for detecting irregularities, such as temperature or speed profiles, which provide information on the failure or wear of machine parts without ever exceeding limit values.
Anomaly detection brings a safety advantage and an increase in productivity: new behavior patterns, vibrations, oscillations or barely recognizable operating points are detected in real time and one can react before damage occurs. After the teach-in phase, i.e. the normal behavior of the system, any deviation is detected and displayed as an anomaly, be it a defective sensor or bearing damage.
Operated with the rotary table control EF3 or the W.A.S. Control Packages, WEISS can supply the necessary data for its components and offer its customers an individually developed monitoring concept.