Fusion of integrated multisensor data for tool wear monitoring

This paper introduces the concepts of data fusion and multisensor integration and highlights the inherent suitability of artificial neural networks for such tasks. The problem of on-line tool wear monitoring in turning operations is approached by applying a three layer backpropagation network for fusion of three machining performance indicating features. After training, the network was capable of classifying previously unseen data into five discrete categories of flank wear.