A New Programming Approach for Robot-based Flexible Inspection Systems

Abstract

Nowadays, flexible inspection systems still have to find their way into high-variety manufacturing environments due to the huge effort required to invest in programming activities. This effort is necessitated by the myriad of restrictions that have to be simultaneously fulfilled during digitization operations by both the optical sensor and the robot manipulator. Consequently, the research described in this doctoral dissertation is geared towards developing a task-oriented programming methodology that reduces the amount of labor currently necessary for programming flexible inspection systems. The presented methodology is characterized by a feature-based description of the inspection task, the automatic generation of sensor trajectories, the automated synthetization of feasible movements of the manipulator carrying the sensor and the online adaptation of the scan trajectories at execution time. One of the key advantage of the presented approach is that human operators are shielded from technical details and therefore do not need to familiarize themselves with either robotics or metrology.