The increasing flood of information through measured process parameters often ends in nirvana, without actually allowing any deep insight into causalities. In the 1940s first mathematical models, under the concept of artificial intelligence, attempted to master these problems. Thanks to significantly increased computer performance, this faculty is today showing amazing results, such as artificial neural networks. Machines seem to learn like people do, they understand languages, analyse pictures and beat humans in chess, Go or even poker.
IAV, the world’s third largest engineering service provider, with more than 6,500 employees at 32 locations in 13 countries, is a cooperating partner of KÜNKEL WAGNER using such approaches to help the sand preparation plants and moulding plants of the foundry industry into the digital age. For over a year now, usual process variables have been collected systematically and evaluated at M/s Heinrich Meier foundry in Rahden. The aim is to identify patterns within the measured values in order to allow clear conclusions to be drawn about the prospective mould strength and to detect anomalies. For this purpose among others machine learning methods shall be applied.
By now it has already been possible to make the measured values of a return sand cooler as transparent to predict the values at the cooler discharge based on the constellation of the cooler inlet values.
Stay updated with KÜNKEL WAGNER.