An ML-Based Approach for Optimizing the Productivity and Efficiency of the Apparel Industry by Focusing on Trainee Employees.

In the apparel industry, training is crucial because it brings skilled workers and promotes increased productivity. However, typical manual approaches frequently fail to accelerate the training process, resulting in unsatisfactory results properly. In this paper, the authors describe an innovative strategy to increase the productivity and efficiency of sewing machine operator training processes by using a machine learning-driven web-based application. The proposed application leverages the power of machine learning models to identify and solve crucial areas for improvement. It specifically detects wrong hand movements, incorrect trainee sitting postures, defects in sewed garments, and errors in dexterity tests during the training period of the sewing operators. Notably, the Graphical Neural Network (GNN) model detects erroneous hand movements with an astonishing 85% accuracy. The Convolutional Neural Network (CNN) model excels in detecting incorrect sitting postures, with an impressive 75% accuracy. Furthermore, the CNN model detects garment defects with an accuracy of 95%, while the CNN model detects test result errors in dexterity tests with an astounding 97% test accuracy. By using the proposed web tool for screening, the authors expect to see a significant increase in trainee productivity and efficiency. Lastly, the machine learning-driven web-based application is a great tool for optimizing the garment industry's training process. Future plans include increasing the application's functionality, introducing new features, and investigating its applicability across multiple sectors within garment manufacturing. By adopting this unique approach, the apparel sector can achieve significant gains in training outcomes, resulting in a more skilled and efficient workforce. [ABSTRACT FROM AUTHOR]