Tool life prognostics in CNC turning of AISI 4140 steel using neural network based on computer vision

Abstract

One of the essential requirements for intelligent manufacturing is the low cost and reliable predictions of the tool life during machining. It is crucial to monitor the condition of the cutting tool to achieve cost-effective and high-quality machining. Tool conditioning monitoring (TCM) is essential to determining the remaining useful tool life to assure uninterrupted machining to achieve intelligent manufacturing. The same can be done by direct and indirect tool wear measurement and prediction techniques. In indirect methods, the data is acquired from the sensors resulting in some ambiguity, such as noise, reliability, and complexity. However, in direct methods, the data is available in images resulting in significantly less chances of ambiguity with the proper data acquisition system. The direct methods, which provide higher accuracy than indirect methods, involve collecting images of worn tools at different stages of the machining process to predict the tool life. In this context, a novel tool wear prediction system is proposed to examine the progressive tool wear utilizing the artificial neural network (ANN). Experiments were performed on AISI 4140 steel material under dry cutting conditions with carbide inserts. The cutting speed, feed, depth of cut, and white pixel counts are considered as input parameters for the proposed model, and the flank wear along with remaining tool life is predicted as the output. The worn tool images were captured using an industrial camera during the turning operation at regular intervals. The ANN training set predicts the remaining useful tool life, especially the sigmoid function and rectified linear unit (ReLU) activation function of ANN. The sigmoid function showed an accuracy of 86.5%, and the ReLU function resulted in 93.3% accuracy in predicting tool life. The proposed model’s maximum and minimum root mean square error (RMSE) is 1.437 and 0.871 min. The outcomes showcased the ability of image processing and ANN modeling as the potential approach for developing a low-cost industrial tool condition monitoring system that can measure tool wear and predict tool life in turning operations.

Methodology for Tool Wear Detection in CNC Machines Based on Fusion Flux Current of Motor and Image Workpieces

“…The estimations were validated against SMr2 (valley material portion) values measured with a microscope. Bagga et al [13] employed an artificial neural network (ANN) considering the machining parameters for the turning of AISI 4140 steel and used images captured inside the working area of the carbide tools at specific time intervals. The worn area was obtained by processing the images by filtering, enhancement, thresholding and calculating the number of pixels in the area, and the remaining useful life was determined with two activation functions, sigmoid and rectified linear unit (ReLU), achieving an accuracy of 86.5% and 93.3%, respectively.…”

“…Several literature reviews have identified the main techniques, tools and trends for processing [21][22][23][24]. Firstly, signals are processed directly in the time domain [3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18] or techniques or transforms are used for their analyses in the frequency [3,5,[18][19][20] or time-frequency [19,20] domains. From here, several techniques are used to obtain features that allow the analysis to be carried out in a better way, such as the use of statistical indicators [3][4][5][6]8,14,[18][19][20], time or time-frequency transforms for direct feature extraction [3][4][5]19,20] and, in some cases, methods for the selection of the most appropriate features or dimensionality reduction such as heuristic techniques [5,6,22], linear discriminant analysis (LDA) [19] or principal component analysis (PCA).…”

“…From here, several techniques are used to obtain features that allow the analysis to be carried out in a better way, such as the use of statistical indicators [3][4][5][6]8,14,[18][19][20], time or time-frequency transforms for direct feature extraction [3][4][5]19,20] and, in some cases, methods for the selection of the most appropriate features or dimensionality reduction such as heuristic techniques [5,6,22], linear discriminant analysis (LDA) [19] or principal component analysis (PCA). Subsequently, classification or decision-making techniques tend to use intelligent systems such as different types of neural networks [8,13,14,16,19,25], support vector machines (SVM) [21][22][23][24], hidden Markov models (HMM) [21][22][23][24], fuzzy systems [8,9,17,18] and DL systems [4,11,14,15].…”

“…The use of intelligent techniques is of great value when processing large amounts of information generated by fused sensors, allowing a more adequate manipulation and interpretation of the data. Another point to note is the implementation of the methodologies, since they can be performed online, allowing process monitoring [3][4][5][7][8][9][10][11][12][14][15][16][17][18][19], or offline once the machining of a part has been carried out [6,13,20,26]. This is an important aspect, since in methodologies that use vision systems or images, the machining process must be stopped for the acquisition of images of the tool or they may be taken after the process; however, this depends on the needs and planning of the process.…”

An Overview of the Study of ANN-GA, ANN-PSO, ANFIS-GA, ANFIS-PSO and ANFIS-FCM Predictions Analysis on Tool Wear During Machining Process

“…A flank wear image with ANN was created and used to create a two-step procedure for predicting the life of a tool. Several researchers have considered using ANN to predict tool wear [22][23][24]. ANFIS models were developed to predict the performance of abrasive water jet machining parameters.…”