Tool condition monitoring in deep hole gun drilling: A data-driven approach

Data-driven tool condition monitoring techniques have received attention in manufacturing industry due to their ability to improve effective and efficient decision-making. In this paper, we present a novel data-driven tool condition monitoring method for tool wear estimation in deep hole gun drilling. The proposed method uses the Gaussian process regression (GPR) based on a combination of force, torque, and vibration signal features, which are extracted within a pre-defined segment. The segmentation method is based on the sliding time window approach, to improve the estimation accuracy of the GPR. We also leverage a smoothing method to refine the estimation outputs to reduce noise and outliers. We show the performance of the proposed method using gun drilling experimental data. The results showed that the tool wear estimation accuracy can be enhanced by the proposed method, which considerably outperforms the other methods such as linear regression, ensemble, and support vector regression.