Description of Invention
The national fourth Industrial Revaluation (4IR) policy is the guiding principle for Malaysia to stay ahead of the 4IR curve. IR4.0 is characterized by increasing automation and the employment of smart machines and smart factories, informed data helps to produce goods more efficiently and productively. Keeping this in mind, a smart AI based Tool Condition Monitoring (TCM) System is designed to increase the productivity in factories. Tool wear in turning, is one of the major problems which may lead to production loss and machine down time. An effective tool wear monitoring method is therefore required to minimize the loss. The novelty of the present work is the use of Hilbert Huang Transform (HHT) to extract the instantaneous amplitudes and the frequencies from the tool emitted sound to train a competitive neural network (CNN) that determines the present condition of the cutting tool insert based on its flank wear. HHT is a recently developed signal processing technique more suitable for analyzing non-stationary and nonlinear signals such as tool emitted sound. This AI based smart TCM is expected to produce considerable savings by reducing downtimes for repair and improving the productivity. The overall accuracy of this CNN is 80.56% which is far better than the accuracy, 47.2%, of CNN trained by the features extracted using fast Fourier transform (FFT). Hence HHT is more suitable in designing TCM Systems for smart industries that supports 4IR. Even though, this research focused on only TCM in turning process, this approach can also be extended to smart machine condition monitoring in general, thereby, supporting our Nation’s 4IR policy.