From Simulation to Shop Floor: A Human-Centered Digital Twin Methodology for Industry 5.0

Abstract

Recent advances in Digital Twin technology focus on visualization, operator training, and real-time machine simulation, aligning with the Industry 4.0 paradigm. However, the transition toward Industry 5.0 demands human-centric approaches that integrate workers not merely as observers but as active, monitorable parts of the system. Despite growing research interest, mature end-to-end methodologies for creating and deploying reliable Human Digital Twins (HDTs) in industrial environments are still lacking. This paper introduces Industrial Meta-Human (IMHU), an end-to-end human-centered Digital Twin methodology designed to bridge this gap. By spanning the entire lifecycle, from human modeling to production deployment, IMHU leverages Unreal Engine simulation to generate accurate human models and synthetic data, allowing safe replication of hazardous scenarios without disrupting ongoing operations. The methodology integrates Artificial Intelligence (AI) to enable real-time monitoring and support data-driven decision-making. Deployed on a fully operational production line, IMHU includes a system integration layer based on a Service-Oriented Architecture (SOA), enabling seamless interoperability with legacy Industry 4.0 infrastructures. Experimental results demonstrate the feasibility and confirm the effectiveness of real-time human-state tracking, and underscore its potential to advance scalable, human-centered Digital Twin systems for Industry 5.0.