The transition toward sustainable manufacturing requires not only high-performance production strategies but also transparent and interpretable decision-making tools. While Machine Learning (ML) has been widely applied in Additive Manufacturing (AM) to predict process outcomes, its adoption often suffers from a lack of interpretability. This study addresses this gap by integrating Explainable AI (XAI) into a data-driven framework for Wire Arc Additive Manufacturing (WAAM), using real experimental data from the deposition of Invar 36 alloy. Two ensemble ML algorithms, XGBoost and Random Forest, were employed to predict key output variables, such as layer width, height, specific energy consumption (SEC), and Global Warming Potential (GWP). SHAP (SHapley Additive exPlanations) values were used to interpret model predictions, revealing feature interdependencies and their relative contributions to each target. By coupling predictive accuracy with interpretability, the proposed framework provides actionable insights for the multi-indicator interpretation of WAAM processes, supporting both energy efficiency and environmental sustainability in AM.

The advent of Industry 4.0 has revolutionised manufacturing systems, introducing unprecedented levels of customisation and variability. Traditional methods for controlling Work-In-Progress (WIP) often fall short in these dynamic environments, necessitating the development of adaptive and intelligent control strategies. This paper explores the application of Reinforcement Learning (RL) for adaptive WIP control in semi-heterarchical architectures for flow-shop production systems. We propose a novel framework that integrates RL, specifically Deep Q-Networks (DQN), with Discrete-Event Simulation (DES) to derive optimal control policies without relying on closed-form mathematical models. Preliminary simulation experiments demonstrate the effectiveness of the proposed approach in handling variations in job processing time variability and throughput reference targets, showcasing the merit and potential of RL for adaptive WIP control.