AI Models that can Predict Concrete Deterioration Developed by Sharjah Scientists
Researchers claim new machine learning models offer insights into when reinforced concrete may crack and weaken.
Researchers at the University of Sharjah have developed machine learning models capable of predicting when and why spalling, a common form of concrete deterioration, may occur. This could provide engineers with crucial insights to prevent damage in structures like bridges, buildings, and pavements.
Understanding Spalling in Concrete
Reinforced concrete is widely used in construction due to its strength, but it is susceptible to cracking and spalling—a process where the surface concrete flakes away, often due to corrosion of the steel reinforcement. This weakening can lead to significant safety risks, especially in large infrastructure like bridges and car parks.
The team’s research, published in Scientific Reports, “investigates the multifaceted factors influencing spalling, employing a comprehensive approach that integrates statistical and machine learning techniques for predictive modeling.” The authors used “descriptive statistics [that] meticulously profile the dataset, emphasizing age, thickness, precipitation, temperature, and traffic parameters.”
Statistical and machine learning models for predicting spalling in CRCP Scientific Reports
Key Influences on Deterioration
Among the key factors identified in the study were the age of the concrete, climate variables like temperature and humidity, and traffic volumes, measured by the Annual Average Daily Traffic (AADT). These elements, the researchers found, significantly influence the durability of Continuously Reinforced Concrete Pavement (CRCP), which is commonly used to avoid the need for maintenance-heavy joints in roads.
Professor Ghazi Al-Khateeb, lead author and professor of asphalt pavements at Sharjah University, explained, “The research study identified key factors influencing spalling in Continuously Reinforced Concrete Pavement (CRCP), including age, climate variables such as temperature, precipitation, and humidity, as well as Annual Average Daily Traffic (AADT) and pavement thickness.”
He added, “These factors were highlighted as primary contributors to the deterioration of CRCP.”
Machine Learning in Action
The researchers used Gaussian Process Regression and ensemble tree models, which, according to Al-Khateeb, “demonstrated the highest accuracy in predicting spalling.” The authors employed “a systematic methodology, comprising several stages to ensure robustness… Descriptive statistics meticulously profile the dataset, emphasizing age, thickness, precipitation, temperature, and traffic parameters.”
The study highlights that model performance may vary depending on the characteristics of the dataset, stressing the “importance of careful model selection” for different applications. The authors noted that “in machine learning, adopting models, including Gaussian Process Regression and ensemble tree models, is grounded in their diverse capabilities and suitability for the complex task at hand.”
Future Research and Practical Applications
Looking ahead, the researchers recommend further studies that integrate real-time sensor data to improve predictions. Professor Al-Khateeb suggested that “exploring the temporal changes in spalling patterns, integrating real-time sensor data for more accurate predictions, addressing regional data variability, and examining the impacts of climate change on spalling… could significantly improve predictive methodologies and enhance the resilience of concrete pavement systems.”
For now, the study highlights the importance of incorporating machine learning into maintenance strategies to reduce the risk of spalling and enhance the resilience of concrete infrastructure.
