| Abstract Scope |
Artificial intelligence (AI) is reshaping fusion and solid-state joining, enabling welding processes that are designed, validated, and controlled through data-driven intelligence rather than narrow, empirically tuned windows. High-volume industries such as automotive, shipbuilding, and aerospace must achieve 100% in-line weld quality prediction across heterogeneous welds, diverse materials, and evolving designs. Traditional trial-and-error development, conservative tolerances, and offline inspection cannot meet this challenge without sacrificing agility or cost. This plenary presents a vision in which human expertise and AI collaborate to deliver robust, adaptive, and certifiable welding solutions. Modern AI tools ingest large, multi-modal datasets, learn relationships between inputs and weld quality, and surface real-time insights for process optimization, risk quantification, and rapid procedure development. On the shop floor, AI architecture supports continuous monitoring, predictive quality, and maintenance while enabling highly customized welding. The talk outlines key technical building blocks, deployment lessons, and a roadmap toward AI-centric welding ecosystems. |