Predicting the Sewability of Heat-Resistant Fabrics as An AI Technique.

Document Type : Original Article

Authors

1 PhD student at King Abdulaziz University / Lecturer at um Al-Qura University

2 Professor of clothing manufacturing, Department of Fashion and Textile, King Abdulaziz University

Abstract

Prediction plays an important role in improving quality of life by anticipating risks and problems before they occur and finding solutions to them. Prediction in the garment industry is used to predict the quality of sewability of the material. This research aims to create a model to predict the sewability of heat-resistant fabrics using machine learning and verify the efficiency of this model and then build a program to predict the sewability and evaluate it by specialists.
24 samples of aramid fabrics of two different weights were prepared, using a needle number (80 and 100), and stitch density (10 and 12 stitches/inch) and a thread denier (27.8, 46.2 and 70.8 tex). After several lab tests on the samples, the machine was trained to predict the sewability of heat-resistant fabrics using this data. The efficiency of the prediction model was verified by preparing ten new samples that had not been used in the machine training process. The samples were tested in the lab and the T-Test was then applied to compare the means of independent groups.
The study found that there are no differences between the mean of actual test scores of the samples and the mean of prediction model scores. A program was also built to predict the sewability of heat-resistant fabrics. it was assessed by specialist teaching staff and owners of factories. The mean evaluation scores of the questionnaire were found to be high and there were no statistically significant differences between the results reached by the specialized assessors.

Keywords

Files

Share

How to cite

Statistics