A laboratory experiment to teach some concepts on sensor-based robot assembly systems

A laboratory experiment is described that can be performed by electrical engineering graduate students and senior undergraduates. The experiment involves the assembly of a diesel-engine oil pump with an IBM RS-1 robot. The robot is instrumented with a force-sensing gripper and an optical binary sensor to detect parts in the gripper. A GE Optomation binary vision system is also used to monitor the manipulator workspace through a charge-coupled device (CCD) camera mounted on the robot arm. The students are required to program the manipulator and to process the information from the various sensors, which are used to recognize assembly components and to verify assembly operations in real-time. This assembly experiment is designed to familiarize electrical engineering students with the problems which dominate robot-based assembly systems and to demonstrate a number of practical sensor-based motion strategies which can overcome part and robot positional uncertainty found in typical manufacturing environments. These sensor-based motion strategies improve the reliability of a robot assembly system. The sensor-based assembly experiment, the laboratory setup, and the sensor-based motion strategies are discussed in detail