The Neural Network based Periodic Tool Inspector (N2PTI) is introduced. N2PTI requires periodic cutting operations on a test piece next to the workpiece. The cutting forces are measured while a slot is being cut on the test piece and a neural network estimates the tool life from the variation of the feed and thrust direction cutting forces. The estimation accuracy of three neural networks (backpropagation, two versions of Probabilistic Neural Networks) is compared. The advantages of the N2PTI are its simplicity, low cost, reliability, and simple computational requirements.
