Automatic welding-robot programming based on product-process-resource models

Abstract

This paper describes a novel end-to-end approach for automatic welding-robot programming based on a product-process-resource (PPR) model, for one-of-a-kind manufacturing systems. Traditionally, the information needed to program a welding robot is processed and transferred along the manufacturing organisation’s value chain by using several stand-alone digital systems which require extensive human input and high skill to operate. A PPR model is proposed through this research as a platform for storing and processing the necessary information along the value chain seamlessly. Unlike existing approaches which make use of complex algorithms to automatically identify the weldment seams, the approach suggested in this research makes use of information already digitalised by design engineers under the form of ISO 2553:2019 compliant weldment annotations. Hence, the PPR model contains the weldment annotations; it enables the automatic programming of welding robots and reduces human input down to a few minutes only. The applicability in manufacturing of the theoretical concept is demonstrated through technical implementations tested in the laboratory and on the value chain of an engineering-to-order (ETO) industrial partner involved in the metal fabrication industry. The experiments were conducted by creating several products using the proposed artefact. Experiments show that automatic programming of welding robots can be achieved using PPR models. The conducted experiments showed a reduction of about 80% in human input measured in terms of time, when using the proposed solution. The reduction of the human input can free up skilled labour resource which ETO SMEs can reallocate to other tasks.