PROTOTYPE OF A ROBOT CALIBRATION SYSTEM DEVELOPMENT

FASTMAN, INC. IS REQUESTING A PHASE I GRANT TO FUND PART OF THE DEVE LOPMENT COSTS OF A ROBOT CALIBRATION SYSTEM DESIGNED TO IMPROVE THE ACCURACY AND LONG-TERM REPEATABILITY OF ROBOTS. OUR SYSTEM WILL IMPROVE THE PERFORMANCE OF ONLINE AND OFFLINE PROGRAMMED ROBOTS WHICH ARE USED IN THE GUIDANCE MANUFACTURING PROCESS. THE SYSTEM IS BASED ON A MODELING AND CONTROL TECHNIQUE WHICH IMPROVES THE ACCURACY OF THE ROBOT'S TRAJECTORY AT ITS END POSITION AND ORIENTATION (POSE). A CORRECTION ALFORITHM IS USED TO GENERATE JOINT CORRECTIONS WHICH, WHEN ADDED TO THE NOMINAL JOINT SOLUTIONS DETERMINED BY THE ROBOT'S CONTROLLER, MAKE THE ACHIEVED POSE EQUAL TO THAT DESIRED. THE CORRECTION ALGORITHM REQUIRES THAT A MODEL OF THE ACTUAL ROBOT LINKAGE BE DETERMINED. OUR SYSTEM DETERMINES THAT MODEL USING A SERIES OF ACCURATE POSE MEASUREMENTS AND A CALIBRATION ALGORITHM DEVELOPED BY THE OWNERS OF FASTMAN. THE POSE MEASUREMENTS ARE MADE USING AN INEXPENSIVE POSE SENSOR, ALSO DEVELOPED BY THE OWNERS OF FASTMAN. THE CALIBRATION SYSTEM ALLOWS POSES GENERATED BY TASK PROGRAMS, INCLUDING THOSE GENERATED ONLINE OR OFFLINE, TO BE STORED AS QUALIFIED POSES. THE USE OF QUALIFIED POSES MAKE THESE PROGRAMS INDEPENDENT OF THE ROBOT'S PERFORMANCE AND THE PARTICULAR ROBOT OR ROBOT TYPE USED TO PREFORM THE TASK.