Use of a robot arm in study of the motion perception process cd.

3. The robot

The Max Planck Institute bought a KUKA KR-500 robot arm about two years ago

This kind of robot is widely used in industry. Unfortunately the software attached to it didn’t allow the user to define trajectory of the end effector, indeed one could give only the start and stop position and the program was choosing the fastest way between those two points. However this is exactly what the industry needs, it’s not enough for the experiments run in the Cyberneum. Therefore the MPI crew has to write its own software. The main problem is passing so called singularities. Singularities or singularity positions are such positions of end effector to which matches an infinite number of the arm positions. In addition for every position of end effector eight different configurations of arm are possible, so the software must quickly choose the best one (the one that is not only mathematically but also physically correct) and if needed smoothly switch between the arm configurations. Because of the safety reasons some motion limitations were added – the person sitting on the end effector should for example not hit the ground. What is more, sudden stops when the arm reaches its maximal range should be avoided.

My task is to support the programming team with my knowledge about kinematics and dynamics. The first assignment was to calculate the inverse kinematics – all eight possible configurations of the arm for the given end effector position.

By now the scientists working at MPI are not able to use all six degrees of freedom (Cartesian coordinates x, y, z and Euler angles γ, φ, θ) of KUKA – usually they are using only one or two of them. Hopefully after some improvements of the software they would be able to take advantage of the full range of KUKA and run much more complicated experiments.