Augmented Accuracy A human-machine integrated adaptive fabrication workflow for bamboo construction utilizing computer vision

Abstract

Despite being sustainable, strong and lightweight, naturally grown bamboo poles are currently used in restricted building typologies. This is due to the large tolerances in the built structures, which is caused by the variations in the dimensions and geometry of natural material as well as the manual, uninformed and imprecise assembly methods. In previous work, we introduced an adaptive fabrication method for bamboo structures that can monitor the fabrication process and compensate for deviations between built and designed form. As a proof of concept, the method is suitable for small scale bamboo structures in 2D-or simple 3D configuration. This paper extends the previous method by integrating the adaptation strategies into a cohesive fabrication and assembly workflow for large scale complex bamboo structures. To enable that, a more effective sensor localization method, adaptation algorithm, connection and assembly system, as well as web-based user interface are developed. The effectiveness of the proposed methods is demonstrated through the fabrication of a pavilion scale branching bamboo structure that complies with intended geometric boundary conditions. Even though the material has substantial geometrical variations, the final structure shows small geometric deviations and a successful interface with the prefabricated roof elements. Our work shows how vernacular materials and processes can be digitally augmented in order to reliably produce building structures, hence enabling their usage in modern applications to a larger extent.