The foundry industry, particularly the knock-out and cleaning stations for removing residual moulding sand and gating systems after casting, is among the most exposed to occupational hazards and harmful factors for workers.

The assumptions adopted in the automation project for the cleaning process of large-scale castings indicate a potential reduction in processing time by 30%, alongside a 6% total cost reduction per finished casting.

As part of the cleaning process, it is proposed to integrate 3D scanning of the actual part. Based on machine vision imaging, the scanned model is compared to its digital twin. The resulting numerical excess material, confirmed through texture-based surface imaging (identifying residual moulding compounds), defines the specific areas of the casting targeted for abrasive blasting.

The use of a dedicated robotic arm will significantly simplify the scanning operation. Once the 3D camera is returned to the tool storage station and the blasting head toolpaths are generated, the robotic arm will replace the human operator during the abrasive treatment phase. Initially, the process will be semi-automated and require operator involvement. However, after optimization using artificial intelligence algorithms and the development of a comprehensive comparison database, full integration of the processing chain scanning → 3D model → blasting head will be achieved, eliminating human decision-making from the workflow.

A final textured surface scan of the cleaned part will validate and ensure that the casting meets the required surface quality standards.

The AUTOWIND project, no. 1/Ł-KIT/CŁ/2023, titled “Automation of Production Processes for Wind Tower Components Including Recycling and Post-Production Waste Management Technologies”, aims at designing and implementing a fully automated line for the cleaning of large-scale castings.