Relation-based clustering of particle measurements for industrial automation

The objective of this work is to automate intelligently the analysis process of machine particle measurements using a relation-based classification of particle attributes. This is enabled through integration of process information obtained through an image processing system, with an evolving knowledge database for improving the accuracy and predictability of particle analysis. This is achieved by measuring relationships among corresponding attributes from various measurements of the particle. Finally, visualisation technique is proposed that helps the viewer in understanding and utilising these relationships, which enable accurate diagnostics to predict future wear modes.