Progess in Visual Feedback for Robot Arc-Welding of Thin Sheet Steel

Abstract

Using conventional robots for MIG (Metal/Inert-Gas) arc welding on thin (1 to 2 mm) sheet steel pressings is restricted by the difficulty of maintaining accurate fit-up and fixturing. Dimensional variations are introduced by wear of tools and fixtures, effects of variable springback in sheet steel pressings, and thermal distortion during welding. These variations cause errors in the position of the welding torch with respect to the seam of up to ± 3.0 mm, which prevents the formation of good welds. This paper gives a very brief description of our first implemented system that forms and uses semantic models for the automatic visually-guided control of a MIG welding robot. The system uses its models to detect the following dimensional variations in thin sheet steel assemblies: gap width, stand-off error, and lateral error. Lap-, T-, and butt-joints are dealt with. Due to the semanticallybased visual feedback technique reported in this paper, errors in the position of the welding torch are corrected to less than ± 0.3 mm, which is a suitable tolerance for producing good welds in such pressings. Furthermore, gap widths are detected for the purpose of dynamically altering welding parameters.

Presented at the 2nd International Conference on Robot Vision and Sensory Controls—Stuttgart, 2–4 November, 1982

GEC Fellows in Robotics at St Cross College, University of Oxford. The Robot Welding Project is supported jointly by the Science and Engineering Research Council, and BL Technology Ltd.