AS MUCH AS WE MAY HATE TO ADMIT IT, machines can perform highly repetitive tasks better than humans. Worker fatigue on assembly lines can result in reduced performance, and cause challenges in maintaining product quality; an employee who has been performing a repetitive inspection task may eventually fail to recognize a defective product, for example.

Systems analyst Steve Robertson is helping develop an integrated color vision system to improve food processing.

But automating many of the tasks in the food processing industry helps improve productivity and product quality -- and sensor technology developed at Georgia Tech will help the food industry employ more automated processes.

GTRI senior research engineer Wayne Daley and colleagues are using their experience in computer vision technologies to analyze individual companies' sensor needs. Sensors, an important part of intelligent automation, include electronic eyes, ears, noses and similar devices that monitor and control an operation. A computer vision system, for example, can tirelessly watch passing products, accurately identifying and removing items of unacceptable quality before they are shipped to customers. Similarly, an X-ray sensor can scan passing products for foreign matter invisible to the naked eye and reject any contaminated product.

After identifying the specific needs of the food industry, the sensor research team, in collaboration with two local equipment manufacturing companies, began developing a low- cost prototype for performing visual grading tasks using color. This could be useful in the grading and sorting of fresh fruit or to determine the "doneness" of baked goods. The prototype is a color-integrated vision system for quality process applications.

Visual quality control is an important tool in many types of food processing: Color can indicate a defect that makes a product unsuitable for consumption. Much of this color identification is currently performed by humans, but a GTRI sensor system that accurately measures color information promises to significantly increase the application of automatic quality control in food processing.

The sensor research team has also filed for a Record of Invention on a sensor system that tracks products on a belt conveyor. Proper product handling requires that the equipment knows the shape, orientation and position parameters of an object. This device gives an approximation to these parameters coarser than would be obtained with a vision system, but sufficient for many food processing operations, at less than a tenth of the cost. Once the sensor system knows the parameters of the product -- a package of peanuts, for example -- it provides guidance and control information for packaging or other kinds of material handling equipment.
-- Dara O'Neil

Further information is available from Wayne Daley, Electro-optics, Environment and Materials Laboratory, Georgia Tech Research Institute, Georgia Institute of Technology, Atlanta, GA 30332-0823. (Telephone: 404/894-3412) (E-mail: wayne.daley@gtri.gatech.edu)

Table of Contents

Send all questions and comments to Webmaster@gtri.gatech.edu

Last updated: 12 Sept. 1996