Traditionally, a concrete mixture is designed in a laboratory by preparing trial batches to meet the design requirements. This method suffers from exponential increases in the required number of samples and experiments when many mixture objectives (UCS, cost, permeability, density, etc.) need to be optimized or many influencing variables, such as water-to-cement ratio, different types of fine and coarse aggregates, superplasticizer, and a variety of supplementary cementitious materials, are considered in the optimization. Furthermore, the variability of the curing environment (humidity and temperature) and constituent properties (the size and shape of aggregate and type of cementitious materials) may incur marginal differences in concrete performance even with the same mixture proportion.

Concrete mixture optimization is motivated by an ever-increasing need for designers to create concrete mixtures that satisfy multiple – oftentimes competing – performance requirements, including cost, physical properties, mechanical properties, etc. Satisfying multiple competing objectives with a large number of influencing variables is time-consuming and costly, or even impossible using laboratory-based experiments. Hence, machine learning models and optimization algorithms can be applied to address this problem. The complex relationship between concrete properties and its components can be accurately constructed by machine learning models. Meanwhile, the optimization algorithms can search for optimal component combinations to achieve the best performance of concrete.

The purpose of this Special Issue is to publish original research and review articles that cover various aspects of the design and optimization of concrete mixtures using machine learning approaches. Research that considers many-objective mixture optimization for multi-functional concrete is particularly welcome.

Potential topics include but are not limited to the following:

• Modelling concrete properties as a function of its components using advanced machine learning techniques, including data pre-processing, regression, classification, clustering, dimensionality reduction, ensemble methods, and deep learning
• Improving the generalizability of existing machine learning models using large-scale data of construction and building materials
• Designing intelligent systems incorporating machine learning algorithms for automatic real-world design of concrete materials
• Optimizing many objectives of concrete, such as strength, cost, slump, density, etc. using machine learning and optimization algorithms (such as metaheuristic algorithm)
• Developing new machine learning algorithms and metaheuristic algorithms for concrete mixture design
• Discovering new cementitious materials using machine learning techniques