In the field of building materials, the content of natural radionuclides plays an important role according to new German radiation protection legislation.

For our customer from the refractory industry, the correlation between radiological and chemical data with the so-called activity concentration index (in short: index), which quantifies the radiological hazard potential, was investigated using Machine Learning. As a result, a prediction model could be developed which, based on a few radiological analyses using the existing data on the chemical composition of the products, provides a reliable quantification of the index.

Particularly advantageous for our customer are the limitation of the number of complex radiological analyses as well as the possibility of being able to provide suitable information on radiological questions from the clients customers at short notice. These points have a high added value for our customer.