Revolutionizing Slab Casting: Unveiling the Power of AI and Computer Vision

Abstract

In the aluminum industry, slab casting is a critical step to ensure high-quality products. Despite this importance, the industry lacks a reliable and accurate system capable of tracking how the process is evolving. GAP Engineering SA has the ambition to give eyes to the operators, by using cameras and images to monitor the ingot during the casting process and to give them insights about the process by using the power of Computer Vision (CV), Machine Learning (ML), and Artificial Intelligence (AI). On the business side, this development intends to bring novelty to the industry, by proposing a new product to the customers. In the first stage, our solution tracks the dynamic evolution of the butt curl in real time, eliminating the need for costly and invasive sensors. By harnessing the potential of AI to standard cameras and taking advantage of production data, our innovation predicts the ingot's future development with precision. Process managers gain unparalleled control, foreseeing potential ingot failures and avoiding wasting time, resources, and energy. With the help of a Digital Twin, the system can be used to finely tune the casting parameters to achieve the desired product quality. The HAWKEYE project is currently being developed and tested in collaboration with Aluminum Duffel BV (AD). Thanks to the innovative vision of the partners and the equipment already in place on the G6 casting machine, we can use the data presented by the casting pit to refine our models.