Predictive and preventive machine maintenance have become indispensable parts of Industry 4.0. Machine condition monitoring (MCM) systems upgrade consumable and damage replacement practices, as well as periodic maintenance activities, to preventive maintenance or even predictive maintenance. This prevents situations in which machines are halted without warning for parts replacement; which causes additional overhead.
The MCM system visualizes machine status, enabling instantaneous monitor over the life of key components in the machine. Massive volumes of data are recorded to analyze and optimize the production line and minimize machine shutdowns, hence maximizing production and simultaneously increasing machine safety and reducing the cost of equipment maintenance.
Machine Condition Monitoring
To enable preventive maintenance for machines, a series of processes and analyses must be carried out that convert collected sensor data to useful information that predicts the health condition of machines and crucial elements. Sensor generated signals undergo the following processes in the machine condition monitoring (MCM) system:
Data Acquisition: defines trigger-capture conditions, number of channels, amplification, sampling rate and conversion of physical phenomena.
Time Domain Processing: processes the captured raw data and optimizes signal qualities using low consumption methods.
Frequency Domain Processing: an optional processing step that converts time domain data to frequency domain data form fine optimization of signal qualities.
Feature Extraction: extracts meaningful feature messages from a large segment continuous time or frequency domain data through default processing functions.
Interpretation and Output: interprets extracted features and makes decisions.
Communication: sends out feature and decision messages using network communication protocols or digital/analog output signals to connect with control desks or other control modules.