Data-informed Digital Twin for large-scale 3D printing in construction

Large-scale 3D printing promises major benefits for the architecture, engineering and construction (AEC) industry but faces challenges including variable material behaviour, multi-machine coordination and dynamic process control. This paper presents a data-driven digital twin that couples real-time monitoring, predictive modelling and adaptive feedback. Machine parameters are continuously linked to material rheology and print outcomes, forming a virtual representation of the process. A clustering-based analysis classifies material mixtures and drives feedback control of printing parameters, improving stability, accuracy and efficiency. The digital twin is demonstrated on a large-scale setup with two machines operating in parallel and five services forming a closed feedback loop. Experiments show reduced material consumption by 7.5% and more consistent, higher-quality prints when using the predictive digital twin. These results indicate that integrating digital twins into large-scale 3D printing can support more robust, adaptive and scalable production. • Describes a data-driven digital twin framework for large-scale 3D printing in AEC. • Real-time monitoring, predictive modelling and feedback enhance process reliability. • Links machine, material, and printed properties for better process control. • Clustering-based prediction links mixture types to optimal printing parameters. • Experimental validation demonstrates improved print accuracy and process stability.