In the NNATT project model research and experimental work is conducted to identify tunnel- and excavated material with sensor based classification and deep learning. Representative tunnel- and excavation material from Austria is sampled, mineralogically, chemically and geotechnically characterized, sensor based measured in preliminary tests and finally applied in a conceptual pilot plant for material classification at the Zentrum am Berg in Eisenerz. Additionally, the project focuses on opportunities for application in alternative building materials, resulting in saving of primary raw materials and the associated reduction of CO2 emissions.

In 2021, excavated materials, such as tunnel excavation, accounted for around 46.1 million tons, or 60% of Austria’s total waste. Our project proposes an innovative solution for the recycling of excavated materials from tunneling and construction projects by using spectral imaging technology data in deep neural networks to predict rocks and their recycling potentials. By implementing real-time material identification on a conveyor belt based on deep neural networks, we aim to feed the excavated material into a recycling chain. This cutting-edge technology enables us to identify the resources potential of the material, facilitating efficient processing and sorting both on-site and off-site.

The objectives of our project are fourfold: to conserve valuable resources in Austria by maximizing material reuse, to reduce the burden on landfills by minimizing waste disposal, to shorten transportation routes and decrease CO2 emissions associated with material transport, and to promote sustainable practices within the construction industry.

Through the integration of advanced technology and a commitment to sustainability, our project represents a significant step towards creating a future in which excavated material is considered a valuable resource that contributes to a circular economy.