Monitoring Flour Performance in Bread Making

An approach, based on multivariate control chart (MCC) methodology, was proposed by Li Vigni et al., that allows monitoring of flour quality and early identification of flour batches potentially leading to poor performance in production. Using this approach, all rheological properties of incoming flour batches are evaluated multivariately, and these values are projected on a model based on historical data, thus highlighting potential deviances from optimal flour batches employed in the past. This chapter extends this strategy to a more general framework that considers routine flour quality control at the miller and routine control of incoming flour batches at the bakery. This approach offers an interesting tool to detect anomalous flour batches; however, the relationship between technological parameters and bread properties is often poorly known. Multivariate evaluation of bread quality allows one to obtain a more compact and complete representation of production performance than considering univariate control charts for each property separately. Moreover, it allows a more realistic evaluation of product and departure from standards taking into account all different properties simultaneously. Evaluation of the rheological properties of incoming flour batches with an MCC approach helps in assessing the similarities and differences among new deliveries and historical data at a very preliminary step of the production chain so that rational planning of the best recipe to apply to exploit flour properties can be done at the beginning of production, instead of modifying it on the basis of the previous production outcome.