Like other major sectors, the ceramics manufacturing industry is also undergoing its digital transformation. The ceramics factory is a uniquely challenging environment for data-driven AI applications. With furnaces scorching away at thousands of degrees, and sensors and equipment being covered by a continuous accumulation of dust, the ceramics production line gives new meaning to the term "noisy data".
In this environment, losses from quality defects make up a very significant portion of production costs, and to date have been a grudgingly accepted part of business. Defects in each tile vary considerably in nature, and could be caused by any one of dozens of machines and hundreds of parameters which take part in the manufacturing process.
With these challenges in mind, we have partnered with Kale Ceramics, an industry leader and the 12th largest manufacturer of ceramic tiles worldwide. We bring in robust modeling tools, and deep learning machinery which can succeed in exactly this kind of high dimensional, noisy data setting. Our ongoing partnership aims to cut costs from quality defects by up to 50%, by first predicting, then diagnosing and finally preventing them through real-time model-driven interventions.