Method and apparatus for machine monitoring with continuous improvement of a predictive maintenance database

Abstract

An apparatus continually monitors predictive maintenance information and analyzes incoming measurements resulting in recommendations for improving setup information, such as machine information and measurement configurations. Smart sensors generate sensor signals corresponding to the parameters of a machine and a transducer converts the sensor signals into digital sensor data, which is stored into memory. An analyzer determines current operating characteristics of each machine and runs an improvement cycle in which it calculates new setup information. For example, the analyzer may calculate a new alert limit, which is a new measurement configuration that may be saved in the database as a new stored measurement configuration replacing an old measurement configuration. The analyzer may also calculate new machine information. For example, the analyzer may identify features in a frequency spectrum that are characteristic of a particular geartrain that is different from the geartrain identity provided in the predictive maintenance database. The analyzer may automatically change the identity of the geartrain in the predictive maintenance database, or it may suggest the change to the operator. The analyzer also runs an analysis cycle during which current operating conditions of each machine are determined and signals are issued based on the current operating conditions.