Machine Condition Monitoring (MCM) is one of the most important strategies in the Industrial Internet of Things (IIoT). MCM also has many pitfalls, even with the computer power and network capabilities we enjoy today. Continuous Vibration monitoring, will very quickly lead to many large data sets. Large data analytics is the process of examining large and varied data sets, to uncover hidden patterns, failure patterns and correlations, machinery health metrics and other useful information that can help maintenance organizations make more-informed machine condition based business decisions. Vibration data collection methods, database strategies and edge computing to maintain large, complete data sets will be discussed. Computing power applied at the edge, (at the point of data collection) to characterize the change of the vibration levels and identify patterns, then quantify that change and turn this data into actionable maintenance management information. This presentation will explore the application of Large Data Analytics through the three types of vibration data collection; periodic data collection, semi continuous data collection and continuously acquired data. The goal of Large data analytics is to perform analysis on the current data and previous data sets to identify the point where machine failure began. The Large Data model necessitates the ability to analyze previous data with the same focus as the current data. The focus of the analysis will shift from simply answering the question of when will the machine fail, to the question of what caused or why or how did the machine begin to fail. This effort and process will naturally lead to identifying the real root cause of failure. From this vantage point repairs and corrections can be implemented that will increase the long term reliability of the machine.
“R.A.I.” the Reliability.aiTMChatbot
You can ask "R.A.I." anything about maintenance, reliability, and asset management.