AI-Driven Integrated Robotic Systems for Smart Material Handling and Real-Time Inspection in Fabric Manufacturing

The cloth manufacturing industry relies heavily on effective defect detection and quality assurance to maintain product standards and reduce material waste. Traditional fabric inspection methods, whether man- ual or rule-based computer vision approaches, often fall short in handling various fabric patterns, lighting variations, and real-time performance demands. This paper proposes an integrated system combining YOLO- based object detection models with robotic arms to automate both defect detection and material handling. YOLO’s fast, accurate defect localization capabilities are paired with robotic manipulation, enabling precise, real-time responses such as marking, removing, or diverting defective sections of fabric. Enhancements like attention mechanisms, multispectral imaging, and hybrid CNN-YOLO architectures improve performance on complex textures, ensuring higher detection accuracy. The proposed system significantly reduces manual intervention, boosts production throughput, and adapts to different fabric types with minimal reconfiguration. This paper also discusses the challenges of dataset limitations, hardware constraints, and synchronization complications, while outlining future directions involving multi-axis robotics, lightweight AI models, and edgebased deployments for smarter, more flexible manufacturing environments.