Predictive maintenance systems and methods of a manufacturing environment

A system for monitoring one or more manufacturing components, the system comprising:

• a machine learning system comprising an unsupervised learning system, a supervised learning system, and a recurrent neural network, the machine learning system comprising one or more processors;
• the machine learning system comprising one or more nontransitory computer-readable mediums including instructions that are executable by the processor, wherein the instructions include:
• obtaining, by the unsupervised learning system, vibration data associated with the one or more manufacturing components and auxiliary data associated with the one or more manufacturing components, wherein the auxiliary data associated with the one or more manufacturing components includes temperature data, electrical data, pressure data, flow rate data, cycle count data, input-output data, degradation data, or a combination thereof;
• clustering, by the unsupervised learning system, the vibration data to generate a plurality of clusters;
• determining, by the unsupervised learning system, a vibration characteristic of the one or more manufacturing components based on the plurality of clusters;
• comparing, by the supervised learning system, the auxiliary data and an auxiliary data prediction model associated with the one or more manufacturing components;
• determining, by the supervised learning system, an auxiliary characteristic of the one or more manufacturing components based on the comparison of the auxiliary data and the auxiliary data prediction model;
• determining, by the recurrent neural network, a state of the one or more manufacturing components based on the vibration characteristic and the auxiliary characteristic;
• generating, by the recurrent neural network, a predictive health model for controlling the one or more manufacturing components based on the vibration characteristic and the auxiliary characteristic, wherein the predictive health model correlates the vibration characteristic and the auxiliary characteristic;
• and
• broadcasting, by the recurrent neural network, a notification based on at least one of the state and the predictive health model.

A method of monitoring one or more manufacturing components, the method comprising:

• clustering, by an unsupervised learning system, vibration data associated with the one or more manufacturing components to generate a plurality of clusters, wherein the vibration data associated with the one or more manufacturing components includes phase data, coherence data, frequency data, amplitude data, or a combination thereof, and wherein the plurality of clusters are generated based on one of a connectivity-based clustering routine, a self-organizing map clustering routine, a centroid-based clustering routine, a density-based clustering routine, and a distribution-based clustering routine;
• determining, by the unsupervised learning system, a vibration characteristic of the one or more manufacturing components based on the plurality of clusters;
• comparing, by a supervised learning system, auxiliary data associated with the one or more manufacturing components and an auxiliary data prediction model associated with the one or more manufacturing components, wherein the auxiliary data associated with the one or more manufacturing components includes temperature data, electrical data, pressure data, flow rate data, cycle count data, input-output data, degradation data, or a combination thereof;
• determining, by the supervised learning system, an auxiliary characteristic of the one or more manufacturing components based on the comparison of the auxiliary data and the auxiliary data prediction model;
• determining, by a recurrent neural network, a state of the one or more manufacturing components based on the vibration characteristic and the auxiliary characteristic;
• generating, by the recurrent neural network, a predictive health model for controlling the one or more manufacturing components based on the vibration characteristic and the auxiliary characteristic, wherein the predictive health model correlates the vibration characteristic and the auxiliary characteristic;
• and
• broadcasting a notification based on at least one of the state and the predictive health model.