Sensory data mapping and database development for interlayer bond strength: Towards digital shadows in 3D concrete printing

Quality control in digital fabrication with concrete can be advanced through a more holistic quality assessment strategy that emphasizes process monitoring during fabrication. Conventional approaches are based on destructive testing, which is resource-intensive, slow, and unable to capture process variations. To overcome these limitations, this study combines in-line, non-destructive sensory data with specimen-based destructive tests, generating a dataset in which production conditions and mechanical performance are directly linked. A dedicated data fusion methodology ensures that measurements from the multi-sensor environment in the 3D concrete printing (3DCP) facility are accurately mapped onto the digital representation of the printed object, regardless of sensor location. The result is a cohesive, autonomous digital shadow in which sensory data is co-registered with experimentally measured mechanical properties. A large-scale experimental program was conducted that produced more than one thousand specimens under systematically varied process parameters. Although the primary focus was on interlayer bond strength, compressive strength and other mechanical properties were also included. The resulting open-access database comprises more than 46 sensory features, each containing over fifty thousand data points. The paper concludes with an exploratory analysis illustrating data variability and univariate relationships; multivariate predictive modeling of mechanical properties is beyond the present scope. By embedding quality assessment into the production phase, this work lays the foundation for adaptive feedback control of interlayer bond strength in 3DCP.