System for predicting the failure probability of imaging devices

Abstract

System for predicting the probability of failure of an imaging device, comprising: • a data acquisition module configured to receive at least one raw data set from the imaging device in real time; • a preprocessing module configured to transform at least one raw data set into at least one structured data set; • a data storage module configured to store at least one structured data set in a time series database; • a predictive analysis module configured to determine at least one predictive outcome for the probability of failure of at least one component of the imaging device, wherein the predictive analysis module comprises: ◯ a first sub-module for machine learning that is configured to calculate an initial risk score for the probability of failure within an initial period using an initial supervised learning model; ◯ a second machine learning sub-module configured to calculate a second risk score for the probability of failure within a second period that is longer than the first period, using a second supervised learning model; ◯ a regression submodule configured to calculate a time-to-failure score for at least one component using a regression model; ◯ a sub-module for statistical analysis configured to detect at least one anomaly in the imaging device using one or more statistical measures; • an output interface configured to output one or more of the first risk scores, the second risk scores, the time-to-failure score, or the detected anomaly; • a connectivity device that is operationally connected to the data acquisition module and configured to transmit the raw and structured technical data over a network, with the connectivity module being implemented as a server, IIoT device or hybrid configuration.

Images