Spectrim is a modular and upgradeable optical sorting platform designed to detect external defects and classify items. Advanced machine-learning algorithms allows for a high level of automation and eliminate variations in grading outcomes caused by inconsistent approaches associated with differing human operators, Compac said.
The system captures up to 500 HD images of a single piece of fruit as it passes through the machine at a rate of 12 pieces of fruit per second.
“With Spectrim we are able to run without manually sorting — something we've never been able to do before, giving us a huge competitive edge,” said Mikey Hanks, system operator at Washington. “We packed Braeburns for the first time this week and started without any [human] sorters.”