A DATA-DRIVEN DECISION SUPPORT SYSTEM FOR PREDICTING AND PREVENTING WARP YARN BREAKAGE

Abstract

Warp yarn breakage is a major disruptor in weaving, causing costly downtime and quality issues. This paper presents a data-driven Decision Support System (DSS) designed to predict and prevent warp breaks. The system continuously collects real-time data on yarn tension, vibration, and ambient conditions (humidity, temperature). Using a ensemble machine learning model, it analyzes this data to identify patterns preceding a break and generates proactive alerts. Implemented in an industrial setting, the DSS successfully predicted 85% of breaks with a 5-minute lead time, allowing for preventive intervention. This resulted in a 30% reduction in unplanned stoppages and a 7% increase in production output, demonstrating the power of predictive analytics in weaving optimization.