Unlike robots, humans have no problem working hand in hand. However, when robots try to perform tasks that require precision, they always face a major problem: they are not precise enough in combination. This problem is especially relevant when the application requires high precision. One of the main reasons for this is that the robots usually do not know each other’s position or have an absolute calibration of the common coordinate system with the required accuracy. We will present our approach to introduce precise alignment procedures in a robotic cell using optically assisted methods. Subsequent analysis of the acquired data by tailored optimization algorithms provide an accurate and absolute coordinate system for a robot ensemble. The accuracy is typically limited by the workspace environment. Mostly acoustic vibrations will define the lower limit of the absolute precision to retrieve a global coordinate system within the working environment. In this work we present an analysis of common limitations as well as an algorithmic procedure to retrieve a global orthonormal basis for the robotic workcell, independent of the number of robots in the ensemble. As an intermediate result of our ongoing research, we can demonstrate a repeatable adjustment accuracy of less than 100μm euclidean distance from a common center.
|