A REAL-TIME, KNOWLEDGE-BASED SYSTEM FOR GRADING CARCASS BEEF

Abstract

ABSTRACT A knowledge-based system for grading carcass beef was developed in the 'C Language Integrated Production System (CLIPS) environment. This carcass beef grading system gets facts about the carcass as input from the meat grader through a computer keyboard or a voice recognition subsystem, and performs recognize-act cycles to deduce the USDA quality and yield grades of the carcass. This system, which was ported on a Compaq Deskpro 386 computer, processes each carcass for its grades in real time, i.e., the final results of grades are given within a fraction of a second after the inputs are completed. This is a much improved system compared to the system previously developed with the Personal Consultant Plus (PC+) expert system shell. The system in PC+ ported on the same computer required more than 10 seconds to process each carcass. This article presents the implementation of the system in CLIPS as it differs in its inferencing mechanism, interfacing with an external input device, and performance, from the previous system developed in PC+.