Abstract
This study compared popular Deep Learning (DL) architectures to classify machining surface roughness using sound and force data. The DL architectures considered in this study include Multi-Layer Perceptron (MLP), Convolution Neural Network (CNN), Long Short-Term Memory (LSTM), and transformer. The classification was performed on the sound and force data generated during machining aluminum sheets for different levels of spindle speed, feed rate, depth of cut, and end-mill diameter, and it was trained on 30 s machining data (10–40 s) of the machining experiments. Since a raw audio waveform is seldom used in DL models, Mel-Spectrogram and Mel Frequency Cepstral Coefficients (MFCCs) audio feature extraction techniques were used in the DL models. The results of DL models were compared for the training–validation accuracy, training epochs, and training parameters of each model. Although the roughness classification by all the DL models was satisfactory (except for CNN with Mel-Spectrogram), the transformer-based modes had the highest training (>96%) and validation accuracies (≈90%). The CNN model with Mel-Spectrogram exhibited the worst training and inference accuracy, which is influenced by limited training data. Confusion matrices were plotted to observe the classification accuracy visually. The confusion matrices showed that the transformer model trained on Mel-Spectrogram and the transformer model trained on MFCCs correctly predicted 366 (or 91.5%) and 371 (or 92.7%) out of 400 test samples. This study also highlights the suitability and superiority of the transformer model for time series sound and force data and over other DL models.
Highlights
The Fourth Industrial Revolution, or Industry 4.0, represents a paradigm shift in the manufacturing industries through the introduction of intelligent manufacturing systems by integrating the Internet of Things (IoT), big data, and artificial intelligence (AI)
In the case of model M2-A, which uses Convolution Neural Network (CNN) architecture with Mel-Spectrogram, the training accuracy steadily increased for the first few epochs, after which the training and validation accuracy and training and validation loss had an erratic pattern, as seen in Figure 12b,c, respectively
It can be seen that the transformer-based Deep Learning (DL) model trained on Mel Frequency Cepstral Coefficients (MFCCs) data had higher validation accuracy than the model trained on the Mel-Spectrogram data
Summary
The Fourth Industrial Revolution, or Industry 4.0, represents a paradigm shift in the manufacturing industries through the introduction of intelligent manufacturing systems by integrating the Internet of Things (IoT), big data, and artificial intelligence (AI). Digitization, intelligence, integration, the generation of engineering knowledge, and connectivity to Man, Machine, Material, Method, and Environment (4M&1E) make up the foundation of a Smart. Milling operation makes use of cutters with specific machining parameters to remove material from the workpiece while obtaining dimensional accuracy and high surface quality. Micromachines 2021, 12, 1484 obtaining dimensional accuracy and high surface quality. Surface roughness is the crucial quality measurement often employed by an off-line method. Machined parts with surface surface roughness within the threshold are accepted, thosethe outside thediscarded. The surface roughness of machined parts is measured through a profiling techThe surface roughness of machined parts is measured through a profiling technique, using nique, using either a stylus or contactless laser-based methods. SMEs make use of either a stylus or contactless laser-based methods.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
