Artificial intelligence methods for constructing wine barrels with a controlled oxygen transmission rate

Abstract

Oxygen is an important factor in the wine aging process, and the oxygen transmission rate (OTR) is the parameter of the wood that reflects its oxygen permeation. OTR has not been considered in the cooperage industry yet; however, recent studies proposed a nondestructive method for estimating the OTR of barrel staves, but an efficient method to combine these staves to build barrels with a desired OTR is needed to implement it in the industry. This article proposes artificial intelligence methods for selecting staves for the construction of barrel heads or bodies with a desired target OTR. Genetic algorithms were used to implement these methods in consideration of the known OTR of the staves and the geometry of the wine barrels. The proposed methods were evaluated in several scenarios: homogenizing the OTR of the actual constructed barrels, constructing low-OTR and high-OTR barrels based on a preclassification of the staves and implementing the proposed method in real cooperage conditions. The results of these experiments suggest the suitability of the proposed methods for their implementation in a cooperage in order to build controlled OTR barrels.