Pests of granary constitute the serious problem while keeping cereal crops. Infecting the entire amount of stored material can follow through the sparseness of delivered goods together with pest. For most dangerous from the ones we rate granary weevil [Sitophilus granarius (L.)], which lives inside of kernel. It causes degradation of quality and the final effect, up to the 5% of losses. One resolution is that we cannot let to describable situation and we have to identify of weevil while we deliver cereal to granary. We can use classification ability of artificial networks. Data set, which we use for creation of neuronal models, was generated on the basis of received empirical data with using SKCS 4100 (Single Kernel Characterization System) device. Analysis of obtained models was carried out determining usefulness of applying them in the process of the identification of appearing of weevil in kernel.