At the AIMSE 2025 Conference, held in Bochum, Germany (18- 19 Nov.) Dr. Ghezal Ahmad Jan Zia from our partner BAM presented the latest Reincarnate results on AI-assisted sustainable material development using the SLAMD (Sequential Learning App for Materials Discovery) platform.
The presentation, titled “Accelerating Sustainable Material Development with SLAMD: Case Studies Using Recycled Materials,” showcased how sequential learning and AI-driven optimization can accelerate the creation of eco-efficient cementitious materials using recycled resources.
SLAMD – introduced as an advanced digital platform developed within Reincarnate, is a tool that integrates:
- Sequential Learning,
- AI-driven optimization,
- Design-space exploration, and
- Iterative laboratory validation,
to significantly reduce experimental workload and accelerate the discovery of sustainable materials.
The methodology, illustrated in the poster, walks through the full workflow from data analysis and curation to design-space creation, model training, proposal generation, and laboratory evaluation. The concept of “Sequential Learning cycles”, a key feature enabling continuous improvement, was emphasized as a core driver for efficient material design.
Two major Reincarnate demonstration cases were featured to highlight SLAMD’s impact:
DEMO 1 – CEMEX: Optimizing Concrete Mixes Across Exposure Classes
In this case study, SLAMD was applied to support CEMEX Poland in developing concrete mixes that satisfy the requirements of five defined exposure classes (XD3, XC4, XC3, XC1, and X0).
Each exposure class is associated with specific boundary conditions, including:
- Strength class,
- Minimum 28-day compressive strength (fc₍28d₎),
- Maximum water–cement ratio, and
- Minimum cement content,
as summarized in the project table shown in the slide.
CEMEX already had practical experience producing these mixes using up to 50% recycled coarse and fine concrete aggregates. Their new objective within Reincarnate was significantly more ambitious: to increase the combined recycled aggregate content to 120% (sum of fine + coarse fractions), reduce production costs, and still fulfill all performance criteria, with compressive strength as the key requirement.
DEMO 2 – Ragn-Sells: Mortar with High Cement Replacement
The second case study focused on sustainable mortar formulations using:
- Recycled concrete fines (RCF)
- Glass powder, and
- Reduced Portland cement content.
The target was to achieve >30% cement replacement while maintaining a compressive strength of ≥30 MPa.
Graphs demonstrated successful progression of substitution levels across six cycles, highlighting SLAMD’s ability to steer experiments toward high-performance, low-carbon formulations.
The presentation highlighted several important outcomes:
- SLAMD can significantly reduce the number of laboratory experiments, providing cost and time savings.
- Sequential learning effectively identifies promising sustainable binder combinations with lower CO₂ footprints.
- The method supports multi-objective optimization (e.g., strength, CO2).
- Results from partners CEMEX and Ragn-Sells demonstrated real-world feasibility and industrial relevance.
The poster also included a breakdown of CO₂ emissions contributions from cement, aggregates, and water, underlining the importance of sustainable binders in reducing overall environmental impact.
AIMSE is a leading European forum for AI and machine learning in materials science, making it an ideal platform to present Reincarnate’s progress. The conference attracted experts from academia, research institutions, and industry, fostering discussions that:
- Strengthened visibility of Reincarnate within the materials AI community,
- Highlighted the project’s role in advancing circular construction materials, and
- Opened opportunities for cross-project collaboration.
Several attendees expressed strong interest in using SLAMD for sustainable material development beyond concrete, demonstrating the broader applicability of the Reincarnate methodology.