The construction industry faces significant environmental challenges, with cement production contributing 8% of global CO2 emissions and 60% of construction waste ending up in landfills. Traditional materials like cement and concrete are energy-intensive and rely on non-renewable resources, while alternatives like waste slag and fly ash present variability in their properties, making large-scale use difficult.

As the demand for more sustainable materials grows, AI-driven materials design is emerging as a key solution. Christoph Dr. Völker, who was as researcher at Bundesanstalt für Materialforschung und -prüfung (BAM) at the time that the tutorial was produced, introduces SLAMD—a tool that leverages artificial intelligence to design high-performance, eco-friendly construction materials. SLAMD has been developed as part of the Reincarnate project, funded by the EU 🇪🇺, while the tutorial is part of our commitment to make the project’s cutting-edge research available to those making a difference in sustainability.

What is SLAMD?

SLAMD (Sustainable Lab Material Design) is an open-source, platform-independent tool that uses AI to optimize the formulation of construction materials. The tool helps users navigate the complexity of material combinations by accelerating the identification of materials that meet specific performance goals such as compressive strength or minimizing the CO2 footprint.

With SLAMD, users can:

  • Create and blend base materials.
  • Set optimization targets for various material properties.
  • Select promising material formulations for lab validation based on AI-driven predictions.

One of the biggest advantages of SLAMD is its ability to significantly speed up the process of material discovery. In a recent case study, SLAMD demonstrated an acceleration of up to 40 times compared to traditional methods. The process works by collecting data from scientific literature and lab experiments, initializing the AI with a random training set, and iterating through short feedback cycles to find optimal materials.

Over 10,000 simulated design runs proved that SLAMD can dramatically reduce the time and resources needed to discover new materials. Compared to random sampling, SLAMD is 10 times more efficient, offering a targeted way to reach material specifications.

Hands-On Approach with SLAMD

SLAMD offers users the ability to input data through its digital lab twin, a feature that allows the creation of materials based on blending ratios and other formulation parameters. The platform then helps users strategically select the best candidates for testing in the lab.

For example, when creating a new material blend, users can specify base materials like cement powder or aggregates, along with important properties such as CO2 footprint and costs. SLAMD then generates multiple formulations and allows for the fine-tuning of parameters, such as water-to-powder ratios and superplasticizer amounts, to create the best-performing mix.

The AI optimization module is a critical feature of SLAMD, enabling users to set targets like compressive strength, cost, or CO2 reduction. The AI then predicts which material formulations are most likely to succeed based on these targets.

In addition to performance targets, SLAMD’s AI can strategically account for prediction uncertainty. This feature allows users to prioritize materials with higher uncertainties in their predictions, helping to identify formulations that could yield even better-than-expected results.

Getting Started with SLAMD

For those interested in trying out SLAMD, the platform is fully accessible and easy to use. Users can explore its features via the demo website, where they can follow the step-by-step user manual and begin creating datasets, blending materials, and optimizing them using the AI-driven tools.

Action Items:

  1. Visit the SLAMD demo website to explore the software.
  2. Create a dataset in the digital lab twin by specifying base materials, blending ratios, and formulation parameters.
  3. Use the optimization module to set performance targets like compressive strength and CO2 reduction, and strategically select promising material formulations for lab testing.

With SLAMD, AI-driven materials design is revolutionizing how we approach the sustainability crisis in construction, offering a path to more efficient, cost-effective, and environmentally responsible solutions.

WATCH THE TUTORIAL