Experimental Workflow Implementation for Automatic Detection of Filament Deviation in 3D Robotic Printing Process

Robotic 3D Concrete Printing (3DCP) is a process of additive manufacturing using building materials. The system that performs 3DCP is a complex system consisting of multiple parts that are independent of each other. However, conventional 3DCP workflows usually lack automatic monitoring of print quality which can be easily affected for various reasons. This paper proposes an integrated workflow of automatic detection of filament deviation in a 3DCP process. The deformation of the filament is adopted as the criterion for print quality evaluation. A Deep Learning-morphology-based filament width estimation method is developed, and a filament deviation detection algorithm with presence of parametric uncertainties is proposed. This workflow allows to detect width deviations in the printed filament by considering several parameters of the printing system. The integrated workflow is implemented and tested through on-site printing tests.