The production of building materials such as cement and concrete is a major contributor to thegreenhouse gases emitted worldwide, especially of CO2. Further, in Germany alone, the construction industry is responsible for approx. 40 % of the national mineral resources consumption. At the same time, there is only very limited to no ‘real’ recycling in this sector. The current mode of operation so far is rather characterized by down-cycling than recycling: Currently, only 25 percent of the construction waste generated in Germany is recycled - despite scarce raw materials and a shortage of building materials. This is where the joint research project "ReCyCONtrol" comes in, which is coordinated at the Leibniz University Hannover (LUH) and involves several partners from science and industry. The goal is to develop new methods based on the use of artificial intelligence (AI) and computer vision (CV) for a high-precision control of the concrete production process especially with regard to sustainable, resource efficient and climate friendly concretes. In this way, high-quality (re-)use of existing construction demolition material will be made possible. The German Federal Ministry of Education and Research (BMBF) is funding the project with approx. four million euros over the next three years.
Background
There are many reasons for the gap in the building materials cycle, but one main cause can be identified: The composition of the materials to be recycled, the so-called recyclates, is subject to large fluctuations. This has pronounced negative effect on the robustness and quality of the new concretes produced from these recyclates. It is possible to compensate for these fluctuations by significantly increasing the cement content (in some cases by up to 20 percent). However, this is neither economically nor ecologically justifiable. The use of recycled materials, which is desirable for environmental reasons, is thus in direct competition with the required reduction of greenhouse gas emissions in concrete production. Another approach to counteract these fluctuations is the separation of detrimental components by sieving. However, the quality of the separated residual materials is then usually too low to be used even for down-cycling and therefore, have to be deposed. As a consequence, this approach is also not expedient. All of the above named disadvantages of existing techniques (which are basically used worldwide) are ruled out by the ReCyCONtrol-approach, brought forward by the consortium.
Goal
In order to be able to make greater use of concrete recyclates, the ReCyCONtrol research group develops technologies that make it possible to continuously record and control the influence of different compositions of these substances on the end product without negatively affecting the economic efficiency and environmental balance of the end product.
The acronym "ReCyCONtrol" is derived from this objective and represents the central concepts: "REcycling", "CYbernetic", "CONcrete" and "conTROL". The developed techniques however certainly are applicable to all kinds of raw materials. In order to raise concrete production to the Industry 4.0 standard, automated, self-learning process monitoring, control and regulation methods will be required in the future. To detect the fluctuating composition of the concrete raw materials, the project develops AI and CV based methods using contactless imaging sensors, a. o.. On this basis, to control the properties of the end product concrete, additives, that are specially developed for the purpose of this project, will be added during the mixing process.
Figure: For an intelligent control of the concrete production, the mixing process and the aggregate supply are monitored by camera sensors. Intelligent algorithms are used to derive measures in order to control the concrete quality automatically and in real time.
