Increased precision demanded in Industry 4.0 of manufacturing has triggered the necessity to develop automated and intelligent quality inspection of mechanical parts. The traditional manual inspection procedures are not always effective, prone to errors, and cannot be effectively scaled, whereas the traditional deep learning models are not very much interpretable. The present study suggests an Explainable AI-based robotic quality assessment system to evaluate automated mechanical parts with the purpose of overcoming the challenges. The system proposed combines the robotic inspection, the use of vision to detect defects and explainable artificial intelligence into one architecture. A defect detection model based on deep learning is used to detect and localize surface and structural defects, and explainability methods, including Grad-CAM and SHAP, can be used to get interpretable explanations about the model predictions. The robotic addition allows the acquisition of data and inspection in real-time, which is consistent and efficient. The experimental outcomes prove that the proposed structure has a higher accuracy, precision, recall, F1-score, and IoU than traditional models, including CNN, and YOLO. In addition, the use of explainability also brings transparency and confidence in decision-making. The suggested framework provides a scalable, credible and interpretive solution to intelligent quality assessment in the contemporary industrial settings.