ABSTRACT
This chapter describes the concept of real operating data-based models developed under the fourth industrial revolution paradigm. At machine level, the fourth industrial revolution provides for the use of data from different manufacturing assets to gather useful knowledge. Therefore, Industrial Internet of Things (IIoT) technologies are capable of moving away from theory-based or laboratory models to real operating data-based models. If the fingerprint is built using data from only one component, there will be a higher rejection rate, leading to good components being scrapped. Therefore, component fingerprints can be joined to create a cluster, which consensually takes into account the shape of the natural groups from each element. Finding patterns that can help to develop a general fingerprint of the part would be useful for benchmarking the component status prior to installation. Clustering algorithms are highly applicable in this type of analysis requiring the use of a fingerprint behavior pattern as a benchmark, because they are not hard to implement.
