November paper on 'Classification of Noisy Data: An Approach Based on Genetic Algorithms and Voronoi Tessellation' by Abdul Rauf Khan, Henrik Schiøler, Torben Knudsen, Murat Kulahci and Mohamed Zaki
In today’s manufacturing paradigm predictive capabilities in manufacturing can be considered a strategic advantage over competitors. Due to the technological advancement and globalization of the world’s economy, manufacturing industry is going through a transitional phase at an unprecedented pace. Digitalization of manufacturing, the Industrial Internet of Things (IIoT) and the fourth industrial revolution are concepts that are getting monumental attention. The advantageous feature of these concepts is not only the idea of machine to machine (M2M) communication, but as a step forward, the deployment of smart machines in manufacturing i.e., machines that are capable of making more informed and automated decisions. The key to manifesting this idea of informed and automated decision making is an intelligent handling and astute analysis of sensor data. In this paper we are presenting a classification methodology especially designed to deal with the issues related to the sensor data analysis, and failures classification in manufacturing.