Smart Manufacturing Systems (SMSs) have evolved into intelligent, data-driven ecosystems that integrate cyber–physical systems, digital twins, and artificial intelligence to enhance efficiency, sustainability, and resilience. This review synthesises more than 250 recent studies across four domains: manufacturing technologies, systems management, sustainable production, and human–robot collaboration. In process optimisation, hybrid machine learning and genetic algorithms reduce surface roughness in machining by up to 35% and decrease energy use in additive manufacturing by 20–30%. In systems management, digital twins and reinforcement learning enable adaptive scheduling and predictive maintenance, increasing operational flexibility and reducing industrial downtime. Sustainability-oriented research shows that additive manufacturing can cut energy consumption by up to threefold compared with subtractive routes, while aluminium recycling and hot-forming processes lower life-cycle impacts. Furthermore, the integration of ISO 14001, ISO 50001, and ISO 14040 supports consistent environmental and energy performance assessment across sectors. Building on this evidence, the review critically examines recent developments in manufacturing technologies, systems management, sustainable practices, and human–robot collaboration, highlighting emerging paradigms such as explainable AI and human-centric design that strengthen safety, transparency, and resilience. Open challenges and research opportunities are outlined to guide future innovation toward intelligent, adaptive, and sustainable manufacturing systems.

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.

This study quantitatively evaluates the impact of Wire Arc Additive Manufacturing (WAAM) process parameters on the environmental performance of components produced in Invar 36 alloy. An experimental campaign involving 49 parameter sets was carried out by varying wire feed speed, welding voltage, and welding speed. For each condition, electrical signals, shielding gas consumption, and wire usage were measured and converted into parameter-resolved Life Cycle Inventory (LCI) data. A cradle-to-gate Life Cycle Assessment (LCA) was implemented in SimaPro 9.6 using the European CML-IA baseline v3.10 midpoint method, adopting 1 kg of as-built deposited Invar 36 as the functional unit. Results show that feedstock production represents the dominant hotspot (8.68 kg CO₂-eq/kg), while the WAAM stage contributes between 1.13 and 4.12 kg CO₂-eq/kg, leading to a total impact ranging from 9.81 to 12.80 kg CO₂-eq/kg. As a result, this study demonstrates that process parameter selection strongly influences environmental performance. Indeed, Specific Energy Consumption (SEC) ranges from 0.44 to 1.95 kWh/kg, while argon consumption varies between 0.26 and 1.51 kg/kg of deposited material. By analysing the results and excluding unstable or manufacturing-infeasible deposition regimes, the optimal trade-off between process stability and environmental impact is achieved at approximately WFS = 7 m/min, V = 20 V, and WS = 6.5 mm/s. Beyond quantifying the environmental hotspots of Invar 36 WAAM, this study provides a dedicated, parameter-resolved cradle-to-gate LCA based on experimentally measured foreground data collected across 49 process parameter combinations. By combining environmental assessment with feasibility screening of the investigated deposition regimes, the work identifies not only environmentally favourable conditions, but also parameter regions that are technologically viable for WAAM processing of Invar 36. The resulting dataset provides a benchmark foundation for future sustainability-oriented process optimisation and decision support in WAAM.

Wire Arc Additive Manufacturing (WAAM) is increasingly being explored as a viable alternative to conventional subtractive and powder-based manufacturing routes for the production of medium- to large-scale metallic components. This review critically examines recent Life Cycle Assessment (LCA) studies on WAAM with the aim of identifying the conditions, design strategies, and process configurations under which WAAM can offer environmental advantages. Following a systematic review, significant methodological inconsistencies across the literature are identified, particularly in terms of system boundaries, functional units, and impact indicator selection. These inconsistencies include the widespread reliance on aggregated endpoint indicators without corresponding Global Warming Potential (GWP) midpoint results, the use of hybrid or incomplete life cycle inventories, and a limited consideration of how process parameters influence both sustainability outcomes and component quality. Such heterogeneity affects the robustness, comparability, and generalisability of existing studies, limiting the reliability of cross-study assessments and potentially leading to inconsistent interpretations of WAAM environmental performance. Based on these findings, the need for more standardised and transparent assessment frameworks is emphasised, and directions for future research are proposed to improve methodological robustness and support more reliable environmental benchmarking of WAAM.