Digital twin-driven, human-centric ergonomic risk forecasting in automotive and industrial assembly: a collective case study
通过对三家汽车制造商的集体案例研究，提出四种数字孪生模型，从预防到自适应再到个性化预测，帮助减少肌肉骨骼损伤风险，推动以人为本的装配系统设计。
Manufacturing and assembly often involve repetitive tasks, awkward postures, and strict time pressures, increasing the risk of work-related musculoskeletal disorders . Industry 5.0 emphasises human-centred design, where Digital Twins (DTs) can improve safety without reducing productivity. However, current ergonomic approaches remain fragmented, relying on offline analyses or isolated real-time monitoring, with limited predictive, personalised, and adaptive capabilities for human–robot collaboration . This paper presents a collective case study of three automotive manufacturers differing in size, sector, digital maturity, and workstation organisation. Cross-case and maturity analyses revealed four DT models: Model A (simulation-driven prevention), Model B (real-time adaptive ergonomics), Model C (personalised ergonomic forecasting), and Model D (a conceptual ideal integrating adaptivity and personalisation). The models show how DTs evolve from prevention to adaptivity, enabling early risk detection, flexible task adjustment, and inclusive design. While full Industry 5.0 integration is not yet realised, human-centric DTs promise safer, flexible, and ergonomic assembly systems.Practitioner Summary This study shows how DTs can enhance ergonomics in automotive assembly by moving from prevention to adaptive, personalised solutions. The four models identified help practitioners reduce musculoskeletal risks, support flexible task design, and advance human–robot collaboration towards safer, Industry 5.0–ready manufacturing.