In the quest for sustainable construction materials, recycled aggregate concrete (RAC) has emerged as a frontrunner, offering an eco-friendly alternative to traditional concrete. However, predicting the compressive strength (f’c) of RAC has been a persistent challenge due to the variability of recycled materials. Enter Samuel Keown, a researcher from the School of Civil Engineering & Assessment Laboratory at University College Dublin, who has developed a groundbreaking approach to enhance the accuracy of f’c predictions in RAC using hybrid machine learning (ML) techniques.
Keown’s research, published in Advances in Engineering and Intelligence Systems, focuses on the integration of Least Square Support Vector Regression (LSSVR) with two innovative optimizers: the Giant Trevally Optimizer (GTO) and the Dingo Optimization Algorithm (DOA). The goal? To create a robust predictive model that can handle the complexities of RAC and provide reliable f’c estimates.
“The inherent variability in recycled materials makes predicting the compressive strength of RAC a formidable task,” Keown explains. “By leveraging the power of hybrid ML approaches, we can significantly improve the accuracy and reliability of these predictions, paving the way for wider adoption of RAC in the construction industry.”
The study demonstrates that the hybridization of LSSVR with GTO, resulting in the LSGT model, yields the most reliable outcomes. This model achieved the highest R2 value of 0.989 and the lowest RMSE value of 1.618, indicating a substantial improvement over traditional methods. The integration of LSSVR with DOA also showed promising results, further validating the effectiveness of these hybrid approaches.
So, what does this mean for the construction and energy sectors? The ability to accurately predict the compressive strength of RAC can lead to more confident and widespread use of this sustainable material. This, in turn, can reduce the environmental impact of construction projects, particularly in the energy sector, where large-scale infrastructure developments are common.
“As we strive for more sustainable practices in construction, the adoption of RAC can play a crucial role,” Keown notes. “Our research provides a significant step forward in making RAC a viable and reliable option for the industry.”
The implications of Keown’s work extend beyond immediate applications. The synergy between ML and optimization techniques showcased in this research opens up new avenues for solving real-world challenges in civil engineering. As the construction industry continues to evolve, the integration of advanced ML models and optimization algorithms is likely to become a standard practice, driving innovation and sustainability.
For professionals in the construction and energy sectors, staying abreast of these developments is crucial. The ability to predict the performance of sustainable materials with high accuracy can lead to more efficient project planning, reduced costs, and a smaller environmental footprint. As Keown’s research demonstrates, the future of construction lies in the intersection of technology and sustainability, and those who embrace this intersection will be at the forefront of industry advancements.
The research was published in Advances in Engineering and Intelligence Systems, a journal that translates to “Advances in Engineering and Intelligent Systems” in English. This publication underscores the growing importance of intelligent systems in engineering, highlighting the need for continued research and development in this area. As the construction industry looks to the future, the insights gained from Keown’s work will undoubtedly shape the development of more sustainable and efficient building practices.