ANALYZE FACTORS AFFECTING THREAD CONSUMPTION IN A GARMENT AND DEVELOP A MACHINE LEARNING-BASED PREDICTION MODEL

Abstract

ABSTRACT The garment manufacturing industry faces intensified competition, prompting the need for cost control and efficient inventory management. This research addresses the challenges of excess thread stock, leading to increased write-off expenses and environmental concerns. Focusing on predicting sewing thread consumption in underwear fullbrief styles, the study employs statistical and machine learning techniques, considering variables such as garment style, fabric/seam thickness, stitch length, stitch density/SPI, seam type, and estimated wastage. The development of a user interface using Streamlit integrates machine learning models for two types of threads, allowing users to input parameters through an intuitive layout. The user-friendly interface facilitates informed decision-making based on predictions of total thread consumption. The application contributes to reducing write-off expenses, minimizing inventory costs, and aligning with environmental sustainability goals. The research highlights the effectiveness of machine learning models, particularly artificial neural network models, in predicting thread consumption. Overcoming challenges such as overfitting and enhancing generalization, the study emphasizes the need for refining model architectures and exploring additional features. The user interface development emerges as a crucial tool for achieving efficient cost control and sustainability in the garment manufacturing industry. Keywords: Garment Manufacturing, Thread Consumption Prediction, Machine Learning, Artificial Neural Network, User Interface, Streamlit, Cost Control.