Sensor stream mining for tool condition monitoring

In metal cutting processes, as the surface of the cutting tool worn out, it releases certain odorous compounds into the cutting chamber air. This study proposes a novel approach and its associated stream mining methods to monitor the cutting tool condition by taking advantage of this phenomenon. The chemical composition of the gases released during the cutting process can be used to monitor the tool condition very accurately that is not possible with present direct or indirect measurement techniques. The chemical compounds released and captured by an e-nose change as the tool wear progresses. This change can be made significant as appropriate doping material were identified and doped into the different layers of the tool inserts, thus allowing a well trained data mining system to accurately estimate the level of tool wear.