<p>Optimizing the tool life and tool costs through timely replacement to prevent catastrophic tool failure is quite essential in the manufacturing industries. This is achieved by monitoring the tool condition and predicting the tool wear to diagnose premature tool failure. However prediction of tool wear using cutting force signals has always been a challenge, especially in novel process like double tool hard turning. The present work focuses on monitoring the tool condition using cutting force signals and predicting the tool flank wear using ANN for AISI 52,100 bearing steel. The experiments were performed using robust Taguchi L₂₇ orthogonal array design to evaluate the influence of cutting speed, feed and depth of cut on the resultant cutting force and tool flank wear. Mathematical regression models using ANOVA were developed to examine the relationship between the process parameters and the tool flank wear. The experimental values of flank wear for both the tools were utilized for training the neural network model using ANN to predict the tool flank wear. The correlation coefficient (R²) value obtained for the tool flank wear using fresh tool was 98.3% and 98.6% at the front and rear end, 98.7% and 96.6% using dull tool at the front and rear end. For ANN trained model, the R² value obtained was 99.23% and 99.74% using fresh tool at the front and rear end, 99.16% and 97.15% using dull tool at the front and rear end indicating that correlation coefficient (R²) values for tool wear yielded higher prediction accuracy with the ANN model in comparison with the statistical regression model.</p> Graphical Abstract <p></p>