October 16, 2003
Automatic Poultry Inspection System Developed in US Ready for Testing in Commercial Processing Plants
The Automatic Poultry Inspection System developed by Agricultural Research Service (ARS) scientists is ready for its first long-term testing in commercial processing plants, having just successfully passed a four-day test in a commercial broiler-processing plant in the United States.
The recent test was done at speeds of 140-180 birds per minute (bpm), double previous tested speeds--on the fastest line, just after the carcasses are defeathered.
The system has been tested at various commercial processing plants during the past decade.
For the latest test, Yud-Ren Chen, an agricultural engineer at the ARS Instrumentation and Sensing Laboratory in Beltsville, Md., and his colleagues took their inspection equipment to a commercial processing facility. The test was done in cooperation with Stork-Gamco, Inc., of Gainesville, Ga., one of the world's largest chicken-processing plant equipment manufacturers. The Automatic Poultry Inspection System had an accuracy rate of 92-95%, which was verified by a USDA veterinary medical officer.
ARS has a cooperative research and development agreement with Stork-Gamco to commercialize Chen's system and move it into use among the nation's 300-plus poultry processing plants. The system quickly diagnoses all physical or nonmicrobial, biological conditions that allow an inspector to remove a chicken from the processing line.
In Chen's system, when a chicken carcass, on a hook dangling from a moving chain, passes through a light beam, the interruption triggers a scan with a light probe from about an inch away.
A computer using ARS-developed "Automated Poultry Inspector" software to identify variations in external skin color and texture and tissue composition, which are clues to problems, is capable of analysing reflected light.
A second system will be tested in the near future. It uses a digital spectral camera to photograph, with a single click, each chicken at three specially selected light wavelengths.