Insect pests can ruin large quantities of stored agricultural products, resulting in devastating economic loss. Early detection of these pests allows for timely intervention and prevention of further damage to stored commodities prior to consumer complaints. Traditional detection methods can be time-consuming and often require in-person inspection of the grain and/or visual inspection of traps. This can be avoided, however, by using remote sensing devices that detect the presence of pests in stored grain facilities and alert managers to infestations, even when not visible to inspectors. Vibro-acoustic sensors have been shown to detect the sounds and vibrations produced by pest insects in various agricultural products. Here, we assessed the potential for vibrational recording as a means of passive detection of pest insects in stored grain. Using an accelerometer attached to a metal plate and placed on a grain mass, we recorded the vibrations produced by three pest species (Sitophilus oryzae, Rhyzopertha dominica, and Tribolium castaneum) at a low (5 adults/kg), medium (20 adults/kg), and high density (40 adults/kg) of insects. We found significant differences in the number and tempo of vibrations produced by different species, with T. castaneum most easily detected due to their movement near the top of the commodity. Overall vibrational energy increased with insect density regardless of species. These results provide the empirical foundation for further automated detection and classification of stored product pest vibrations using machine learning technology, allowing managers to act quickly to control infestations.
.png)
