In the heart of Nairobi, Kenya, a groundbreaking study led by David Sinkhonde from the Pan African University Institute for Basic Sciences, Technology and Innovation is revolutionizing the way we think about sustainable construction materials. Sinkhonde and his team have successfully demonstrated that waste tire rubber and brick powder can be recycled into high-performance rubberised concrete, a development that could significantly impact the energy sector and beyond.
The research, published in the journal *Applications in Engineering Science* (translated as *Practical Applications in Engineering*), focuses on the compressive strength of rubberised concrete containing brick powder. By employing advanced machine learning (ML) models, the team has paved the way for more accurate predictions of concrete behavior, enabling engineers and practitioners to make informed decisions about mix designs and quality control.
“Concrete is a complex material with variable behavior,” explains Sinkhonde. “Incorporating machine learning models allows us to better understand and predict its performance, leading to more sustainable and efficient construction practices.”
The study explored several ML algorithms, including adaptive boosting (AdaBoost), light gradient boosting machine (LightGBM), extreme gradient boosting (XGBoost), gradient boosting regression (GBR), cluster regression, multilayer perceptron (MLP), and Gaussian process (GP) models. The findings revealed that the GBR model outperformed others in predicting compressive strength, with R2 values ranging from 0.77 to 0.98.
One of the key insights from the research is the identification of age as the most influential variable, with an average SHAP (SHarpley Additive exPlanations) value of 3.561. This was followed by tire rubber aggregate, coarse aggregate, and cement. The study also highlighted the importance of regularization to prevent overfitting, a common challenge in ML models.
The implications of this research are far-reaching, particularly for the energy sector. As the demand for sustainable and energy-efficient buildings grows, the ability to predict and optimize the performance of construction materials becomes increasingly crucial. By leveraging ML models, engineers can design structures that not only meet environmental standards but also reduce energy consumption and costs.
“Our findings open up new possibilities for the construction industry,” says Sinkhonde. “By integrating ML models into the design process, we can create more resilient and sustainable buildings that align with the goals of the energy sector.”
The research also underscores the importance of interdisciplinary collaboration. By bringing together experts from civil engineering, data science, and materials science, the team has developed a comprehensive approach to addressing the challenges of sustainable construction.
As the world continues to grapple with the impacts of climate change, the need for innovative solutions in the construction industry has never been greater. Sinkhonde’s research offers a promising path forward, demonstrating how the integration of waste materials and advanced technologies can drive progress towards a more sustainable future.
In the words of Sinkhonde, “This is just the beginning. The potential for ML models in construction is vast, and we are excited to explore the many possibilities that lie ahead.”