A team of researchers from the University of South Australia (UniSA) has developed a new machine that automates the dual functions of almond quality grading and contamination detection.

In collaboration with industry partner SureNut, the UniSA team created an almond inspection machine that features two high definition cameras, one hyperspectral camera and artificial intelligence (AI) algorithms that autonomously examine almond production lines for signs of damage, double kernels, insect and mold damage, and more.

Source: UniSA

Reportedly, the new system examines almonds in greater detail than traditional grading efforts, which are conducted manually and thus vulnerable to error. Conversely, the hyperspectral camera monitors factors likely invisible to the naked eye such as moisture content, free fatty acid content (FFA) and peroxide value (PV) — which are linked to rancidity — and aflatoxin B1 content, which is a carcinogen.

Additionally, the two high definition cameras examine both sides of the almonds for physical damage such as chipping and other damage.

Although the machine is currently under trial, the researchers report that the automated technique for grading and detecting contamination on almond production lines has improved significantly as has the time it takes to conduct such tasks.