The collaborative efforts between partners Bundesanstalt für Materialforschung und -prüfung (BAM) and the Technical University of Berlin (TUB) have culminated in a groundbreaking preprint paper titled “Beyond Theory: Pioneering AI-Driven Materials Design in the Sustainable Building Material Lab.” Co-authored by researchers Sabine Kruschwitz, Christoph Völker, Elisabeth John, and Rafia Firdous, the preprint not only showcases remarkable results demonstrating the superiority of AI-driven materials design over traditional design of experiment methods but also is expected to support the Technology Readiness Level validation of one of Reincarnate’s digital innovations – SLAMD.

Addressing the challenge of enhancing the sustainability of building materials within complex formulations involving binders, additives, and recycled aggregates aligns with Reincarnate’s goals. In this study, our partners conducted a comparative analysis between Data-Driven Design using SLAMD, the open-source AI materials design tool developed by BAM, and traditional design of experiments methods. The objective was to develop high-performance, waste-based binders surpassing 100 MPa compressive strength after 7 days.

 

Bridging the gap between theoretical potential and practical lab applications

Invaluable insights into the real-world effectiveness of data-driven designs in terms of acceleration of the design process and the quality of the final material designs are some of the many findings of this study. Our partners tested approximately 1,500 samples in the laboratory, validating 80 innovative waste-based cement formulations. 

 

“In the construction industry, we have to push for responsible consumption and production, as outlined in the Sustainable Development Goals. This study is the first time we have demonstrated the capabilities of AI in a real laboratory environment.  The study is a great example of how a paradigm shift in the development of sustainable materials can be achieved in the industry. This milestone presents the necessary conditions for accelerating time-to-market while maximizing sustainability in the construction industry,” expressed Christoph Völker, Data Scientist at BAM.

Building a more environmentally conscious and competitive industry

The integration of AI in material design not only accelerates the development process and enhances the quality and efficiency of the final products but also holds significant potential for market players in the construction industry. The innovation can provide a competitive edge by shortening the time-to-market for new, high-performance materials. 

The faster and more efficient material development facilitated by AI-driven design allows companies to bring environmentally friendly and superior-quality construction materials to the market sooner, positioning them as leaders in sustainable building practices to meet the increasing demand for eco-friendly solutions in the construction sector.

Furthermore, the cost-effectiveness of the AI-driven approach becomes particularly evident in the higher success rate of achieving the target compressive strength. With a success rate of 33% in 8 weeks of development time for AI compared to 4% in 12 weeks of development time for laboratory approaches, the former is faster and more efficient in delivering the desired material properties. This translates to potential cost savings for companies engaged in materials development for applications in the construction industry.

This preprint has been prepared for the upcoming RILEM Spring conference in Milan, Italy, which will take place on April 10-12. This follows BAM’s success at the 4th RILEM conference in Morocco in 2023. By publishing the official paper, they hope to achieve the TRL level for Reincarnate, which will allow them to apply it in real-world scenarios.

Read the full paper here!

 

🔎 Key findings of this research

  •       AI systems indicate a direct path to the design target – they achieved a 33% success rate.
  •       Design of experiments approaches had a low success rate to the design target – only 4% met the design target 
  •       Low-cost – The AI system was more efficient and required only 70% of the initial laboratory budget.

 

Stay tuned to our research developments by subscribing to our newsletter and follow us on LinkedIn and Twitter so you don’t miss a thing!

 

Read our previous articles on research works regarding materials:

circular economy construction materials innovation materials science RILEM Conference SLAMD