This research paper presents a novel method for robotic spraying of glass-fibre reinforced concrete (GFRC) on a permeable reinforcement mesh. In this process, the mesh acts as a functional formwork during the concrete spraying process and as reinforcement once the concrete is cured, with the goal of producing slender reinforced concrete elements efficiently. The proof of concept presented in this paper takes inspiration from ``Ferrocement'' technique, developed in the 1940s by Pier Luigi Nervi (Greco, 1994) and shows how robotic spraying has the potential of producing such slender and bespoke reinforced concrete elements while also having the potential of reducing manual labour, waste and excess material. The system is coined with the name ``Robotic AeroCrete'' (or RAC) in reference to the use of an industrial robotic setup and the pneumatic projection of concrete.

This paper describes the design and fabrication process of a concrete column cast in ultra-thin, 3D printed formwork, using a process known as Eggshell. The column was prefabricated as part of a real-world construction project, serving as the main load-bearing element for a reciprocal timber frame structure. The fabrication of the column required upscaling of the Eggshell process, to allow for the fabrication of elements of an architectural scale. Furthermore, several challenges had to be addressed such as: integration of reinforcement, establishing the formwork design space, and scaling up the 3D printing process. For the production of the final column a 1.5 mm thin formwork was 3D printed, after which it was combined with a prefabricated reinforcement cage and filled with concrete in a set-on-demand casting process. The successful realization of the project provides a first example of a fullscale building element produced with the Eggshell fabrication process. By 3D printing the formwork, geometrical freedom in concrete construction is greatly expanded, as well as formwork waste reduced.

In this paper, we tackle the challenge of detection and accurate digital reconstruction of steel rebar meshes using a set of industrial depth cameras. A construction example under investigation in this paper is robotic concrete spraying, where material is sprayed onto double-curved single layered rebar meshes. Before the spraying process can start, the location and geometry of the rebar mesh needs to be accurately know. We present an automatic image-based processing approach of depth images for grid point extraction at an accuracy of a few mm. Furthermore, we propose a sequence of execution steps in a robotic setup, including the hand–eye calibration, which enables the direct georeferencing of multiple data sets acquired from various poses into a common coordinate system. With the proposed approach we are able to digitally reconstruct a mesh of an unknown geometry in under 10 min with an accuracy better than 5 mm. The digitally reconstructed mesh allows for computation of material needed for its construction, enabling sustainable use of concrete in digital fabrication. The accurately reconstructed digital mesh, generated based on the proposed approach in this paper, is the input for the following spraying step, allowing for generation of accurate spray trajectories.

No abstract

In this paper, we tackle the task of replacing labor intensive and repetitive manual inspection of sprayed concrete elements with a sensor-based and automated alternative. We present a geometric feedback system that is integrated within a robotic setup and includes a set of depth cameras used for acquiring data on sprayed concrete structures, during and after fabrication. The acquired data are analyzed in terms of thickness and surface quality, with both sets of information then used within the adaptive fabrication process. The thickness evaluation is based on the comparison of the as-built state to a previous as-built state or to the design model. The surface quality evaluation is based on the local analysis of 3D geometric and intensity features. These features are used by a random forest classifier trained using data manually labelled by a skilled professional. With this approach, we are able to achieve a prediction accuracy of 87 % or better when distinguishing different surface quality types on flat specimens, and 75 % when applied in a full production setting with wet and non-planar surfaces. The presented approach is a contribution towards in-line material thickness and surface quality inspection within digital fabrication.