Abstract
Costly and time-consuming destructive methods of sampling, curing, and testing under hydraulic jacks are often used to determine concrete properties. Computational intelligence techniques provide the ability to estimate concrete properties quickly at almost no cost. This paper presents a state-of-the-art review of statistical, pattern recognition/machine learning, evolutionary algorithms, and hybrid approaches for estimation of concrete properties such as strength, adhesion, flow, slump, and serviceability using previously collected data. Advantages and disadvantages of the methods are delineated.