Industrial manufacturing relies on increasingly complex and flexible production systems, where variability in processes and machine conditions limits efficiency, quality and predictability. Many critical production decisions still depend on tacit operator knowledge, particularly acoustic cues, which are subjective, non‑repeatable and difficult to scale up. LUCIA aims to transform this hidden expertise into an intelligent and scalable production control solution by integrating non‑invasive acoustic monitoring, edge computing and AI. Advanced AI models will analyse process‑specific acoustic signatures to enable early anomaly detection and predictive decision support. The project will deliver a validated industrial demonstrator integrating real‑time monitoring, production planning and operator interaction, enabling a transition from reactive, experience‑based control to autonomous, explainable and data‑driven manufacturing aligned with SMART priorities.