REMATCH: AI Enables Sustainable Recycling of Excavated Material From Mechanised Tunnelling
Tunnel construction can generate over 200 tonnes of excavated spoil per hour. If this spoil cannot be used on the same construction site, it often ends up in landfill sites. To enable sustainable utilisation, the properties of the material must be known. Herrenknecht, STUVA and TH Köln have now developed the basis for a corresponding AI-supported system.
The collaborative REMATCH project was recently announced as one of the three finalists for the prestigious bauma Innovation Award in the research category. The announcement of the winners and presentation of the award will take place on 6 April 2025 in Munich.
The measuring system consists, among other things, of a ball pendulum, a plow, cameras and lighting equipment above the moving conveyor belt
Quelle/Credit: TH Köln
Newly Developed Instruments Measure Loads in the Material Flow
Depending on its characterisation, excavated material from tunnel construction can be used in a variety of ways, for example as a road base or concrete aggregate. However, the material still needs to be separated as accurately as possible on the construction site. To make this possible, a new measuring system has been developed that is based on artificial intelligence and will determine the relevant geotechnical parameters for sorting in real time.
The relevant parameters include, for example, the so-called settlement rate to describe the workability or flowability of the excavated soils or the shear strength and water content, which have an influence on the stability of the soils to be added. Using the usual methods, these parameters cannot be determined continuously (or only with great difficulty) during operation of an earth pressure balance machine, a widely used type of TBM. For this reason, a ball pendulum and a plough were mounted above the running conveyor belt. The strength with which the passing material moves the instruments is then measured and the load curves are documented. These loads look very different in sandy soil, for example, than in clayey soil. The aim of the research project was therefore to train an artificial intelligence that can deduce the geotechnical properties of the excavated material from the load on the measuring instruments. This in turn enables machine operators to classify the excavated material on the conveyor belt and initiate appropriate sorting.
Tests Under Realistic Conditions
View into an earth pressure balance TBM: The excavated material is transported away at high speed on a conveyor belt
Credit/Quelle: Herrenknecht AG
To obtain a valid database for training the artificial intelligence, the project team first tested various material samples on a circular conveyor belt with a diameter of 2 m at STUVA e.V. in Cologne. To better reflect the conditions on a tunnel boring machine, Herrenknecht AG also set up a test conveyor belt around 50 m long on a 1:1 scale at its headquarters in Schwanau, Baden-Württemberg. There, many different samples could be analysed and classified under realistic conditions. In addition, the new measuring systems developed in the REMATCH project were tested in an actual application in a tunnelling project. This provided a very good basis for determining selected geotechnical parameters of the excavated material and thus the utilisation potential. Further investigations are already being planned.
About the Project
The research project “REMATCH – REsource efficient tunnelling based on real-time excavation MATerial Characterisation” was funded by the German Federal Ministry of Education and Research and the French Agence nationale de la recherche from 2021 to 2024. The project manager on the German side was the Research Association for Tunnels and Transportation Facilities (STUVA e. V.). Other partners alongside the TH Köln were the TBM manufacturer Herrenknecht AG and on the French side the planning and consulting company Arcadis, LIRIS – Laboratory for Image Processing and Information Systems at the University of Lyon, as well as the associated partners DB Netz AG, the French Centre for Tunnel Studies Centre d’Études des Tunnels and the public developer Tunnel Euralpin Lyon–Turin.
