AI-Based Soil Stabilization Using Recycled Concrete for Sustainable Structural Foundations
DOI:
https://doi.org/10.7492/z3237b20Abstract
Weak Soils: The main drivers of foundation stability and construction performance. However, even though conventional soil
stabilization has shown success, it has a negative environmental effect. Thus, sustainable materials with Recycled Concrete Aggregates (RCA),
Artificial Intelligence (AI), are presented as a valid alternative. Soil samples were stabilized using different percentages of Recycled Concrete
Aggregates (RCA) (0%, 10%, 20%, and 30%). In laboratory tests compaction, Unconfined Compressive Strength, California Bearing Ratio,
and Atterberg limits was done. Machine learning models were also used, such as Linear Regression, Random Forest and so on to predict the
soil. Results showed that by incorporating RCA, the strength and bearing ability of soils improved significantly. The most improved
performance in the experiment was obtained by using 20% RCA. Of the AI models, Random Forest outperformed Linear Regression in
prediction quality. All in all, the combined applications of RCA and AI techniques enable a more effective, economical, and sustainable solution
to stabilize the soil with a better applicability in foundation engineering. Keywords: Soil Stabilization, Recycled Concrete, Artificial
Intelligence, Machine Learning, Sustainable Foundations, Bearing Capacity.








