The global welding market is projected to reach USD 34 billion by 2030, exhibiting a compound annual growth rate (CAGR) of 4.7% during the forecast period. This growth is driven by industries such as construction, automotive, aerospace, and energy. Welding processes encounter significant challenges, including material incompatibility, weld defects, and heat distortion. These challenges impact productivity and quality. For instance, high-carbon steels require preheating up to 400°C to prevent cracking, and porosity defects exceeding 3mm can compromise structural integrity. Heat distortion during welding results in rework costs accounting for 5–10% of fabrication expenses. Manual welding productivity lags, with only 50–70% arc-on time compared to 85% in automated systems. Addressing these challenges with artificial intelligence (AI) and advanced monitoring solutions can yield substantial cost savings. AI-powered systems can optimise welding parameters in real-time, predict defects, and enhance productivity. AI-based solutions combined with advanced non-destructive testing (NDT) can improve the efficiency and reliability of weld inspections. The WeldVue 2.0 project aims to implement a framework (based on proprietary development of BUL) that integrates software and hardware. This framework enables process optimisation by monitoring welding processes using advanced NDT systems and approaches (proprietary development by ETherNDE) in aerospace parts fabrication. The project’s objectives include: * Reducing aerospace parts fabrication reconfiguration time by 15% through the development and deployment of an advanced multi-model AI-driven model that optimises and streamlines the manufacturing process. * Establishing an operational (TAI) pilot line consisting of sensors and advanced NDT systems for real-time monitoring and quality inspection of the fabrication process, achieving a defect detection probability of over 99%. * Enhancing quality control measures to reduce scrap and waste by 50%. Process Efficiency Enhancement: An Advanced Software Solution Welding operators currently rely on predefined input parameters to generate specific outputs in welding processes. While existing software solutions can predict output parameters based on input values, they lack the capability to perform the inverse process. Consequently, operators must manually adjust input parameters through trial and error. This project aims to develop an advanced software package that integrates a forward prediction and an inverse optimisation framework. This comprehensive approach enables the identification of optimal inputs required to achieve desired weld characteristics. By automating the optimisation process, manual intervention and trial-and-error methods are eliminated. Manufacturers, such as TAI, will benefit from this software architecture, which facilitates the rapid generation of precise input settings tailored to their specific requirements. Furthermore, a sophisticated in-line monitoring system will be developed, equipped with advanced NDT sensors and artificial intelligence (AI) for continuous welding assessment. This system will detect anomalies or deviations in real time, enabling immediate corrective actions. The closed-loop feedback mechanism implemented in this system enhances process reliability and minimises the risk of defects, ensuring consistently high-quality welds.