LUCAi™ uses artificial intelligence techniques to classify and grade fruit to precise market demands. The classification is derived from a learning set generated by more than a quarter million individually classified images over the last five years. And as part of BBC Technologies’ ongoing Berry Science Program, LUCAi™ will continue to increase its knowledge base as more and more variety-specific and seasonal-specific images become available.
As fruit passes along a grading line, each piece of fruit is photographed by multiple cameras. LUCAi™ then identifies and instructs how each individual berry is to be classified. LUCAi™ can process up to 2,400 individual images of fruit each second. This impressive speed and accuracy are achieved by employing 17,408 graphics processors capable of 304,000,000,000,000 RTX operations per second. LUCAi™ is also able to view fruit in wavelengths not visible to the human eye, further enhancing sorting accuracy.