The term "predictive maintenance" has currently become a buzzword just like other popular concepts such as big data, industrial Internet of things, and smart manufacturing . Everybody is thinking and talking about it - but why?
Predictive maintenance involves key technologies such as machine learning and sensory collection, so the operating data from the machine equipment itself is extremely important. For one it is about the amount and the accuracy of data, and for the other when it was collected (in this case: in real time!).
In regard to maintenance management, it began with post-event maintenance. This was then followed by preventive maintenance and state-based maintenance. Only now it is evolving into the predictive maintenance stage.
Therefore, if we finally reach the end stage of "true" predictive maintenance, we can detect potential failures as early as possible, put forward preventive measures to avoid accidents, and even ensure the safe operation of equipment. Furthermore, we care able to predict future failures and thus guide the development of appropriate maintenance plans. This consequently leads to avoiding excessive maintenance and reducing the costs of equipment maintenance. Information and data about the operating status of the equipment itself is considered to be the basis for achieving advanced maintenance.
Data accumulation
The accumulation of the types and quantities of real-time status data from equipment is critical to the establishment of equipment mechanism models and failure prediction models. It is also considered to be a prerequisite for completing machine self-learning itself.
Condition monitoring refers to the periodic or continuous monitoring of the condition parameters of the running equipment such as vibration, noise, current, temperature, oil quality, etc.. Condition monitoring also includes effective system automatic monitoring analysis of the equipment operating status. Based on the corresponding self-diagnosis status report, maintenance can be guided into the correct way of handling said machinery.
According to statistics, predictive maintenance can increase overall productivity by anywhere between 2% and 40% and reduce maintenance costs by 7%-60%. Equipment life can be extended 1-10 times and spare parts inventory can even be reduced by 10%-60%. Furthermore, it can reduce energy consumption by 5%-15%, and process downtime by as much as 70%.
1. Vibration; 2. Contact Temperature; 3. Relative Humidity; 4. Ambient Pressure
BCM-the "weapon" of data accumulation
First of all, Balluff's condition monitoring sensors can collect data with multi-functional variables, and detect a variety of physical variables, such as vibration, temperature, relative humidity and environmental pressure. Thus, it can greatly reduce the overall system costs, while improving the type and amount of data needed for condition monitoring.
Secondly, BCM can pre-process a large amount of raw data in one device and provide standards for evaluation and analysis. BCM supports a two-way communication through IO-Link, as it transmits the necessary data to the main system, and then easily adjusts the parameters and its use for data evaluation.
With the development of artificial intelligence, big data, cloud computing and the Internet of Things, predictive maintenance based on device condition monitoring will become the new norm. As an expert in innovating automation, Balluff is providing powerful technical assistance in helping companies start and achieve "true" predictive maintenance.
