Production process has been transformed by making complicated geometries possible with less material waste and faster lead times. Although quality assurance continues to face difficulties due to the inherent diversity in additive manufacturing processes and materials. Artificial Intelligence has developed these issues by improving manufacturing quality control through computer vision, machine learning, and multi-sensor data fusion. Therefore, support vector machines improve material categorization, convolutional neural networks improve defects detection, and reinforcement learning improve real-time process optimization. Furthermore, AI applications identify fault localization in photovoltaic systems and surface flaw recognition in steel production to increase the precision of defect detection, streamline production processes and guarantee product quality. Researchers addressed to find out real-world issues such as data preparation, interpretability of models, and smooth integration with current production processes for further development of detecting faults in additive manufacturing. In this research, researchers detected multiple fabric defects by categorizing & classifying them into certain classes by experimenting low data storage requirements & minimal processing times. Therefore, they implemented data cleaning & processing requirements integrated with Gabor Transform for defect classification and achieved 96% accuracy. Researchers also experimented multiple defects identification & classification on automotive industry by applying multiple deep learning approaches. In the future, researchers suggested to work on practical issues finding on this topic including data pretreatment, model interpretability and integration into existed systems.