US Patent 11169868 Automated fault monitoring and management method

A fault monitoring and management method that collects readings from hardware components and software functions to deduce the source of a system failure by utilizing a system representation method based on directed graphs. The presented method utilizes a system description that establishes absolute dependence, which means that the failure of a component leads to certain failure of the successor components that depend on the output of the failed component, between system or process elements. The change in the system behaviour upon the failure of each system element is automatically determined by algorithms that process the graph depiction of the system architecture.