Immersive cognitive factory twin: optimizing industry 5.0 with an integrated VR, ML, and LLM framework

Optimizing modern factories requires dynamic, data-driven decision support aligned with Industry 5.0’s human-centric principles. This research introduces the ‘immersive cognitive factory twin’, a novel framework leveraging virtual reality (VR), machine learning (ML), and large language models (LLMs) to enhance key performance indicators (KPIs). The system features an immersive VR simulation for visualizing factory operations, a neural network trained on simulated data to predict KPIs (e.g. operator success rate), and an integrated LLM to analyze predictions and suggest actionable improvements (e.g. operator reassignments, breaks) via natural language within the VR interface. A proof-of-concept multi-run simulation of a three-station assembly process demonstrated substantial advantages (throughput +28.6%, scrap −35.4%; t-tests and ANOVA, [Formula: see text]). This framework empowers managers with real-time, actionable insights for substantial gains in factory efficiency and resource optimization.