A hybrid tool life prediction scheme in cloud architecture

This paper presents a cloud service scheme to predict tool wear. When lacking sufficient historical tool wear data for building a data-driven model, predicting tool life is challenging while under various cutting conditions with different tools and machines. On the basis of a hybrid tool wear model with dynamic neural network, this paper proposes a tool life prediction scheme for predicting tool wear by given cutting conditions and relevant tool wear features which extracted from sensing segment data. Experimental results show that the proposed scheme can assist factory users to predict various tool lifetimes well in the cloud-service environment while with the first tool samples for modeling.