The LBR I was the first lightweight robot of the Institute of Robotics and Mechatronics. Due to its low weight and the torque sensors integrated in the joints, this robot allowed to investigate future application possibilities of cobots in the field of assembly and human-robot collaboration.
The LBR I was completed in 1995.
Technical specifications
Control Cycles: Current 40 kHz, Joint 2 kHz, Cartesian 1 kHz Electronics Integrated Electronics, external Cabling, Communications by optical SERCOS-Bus
System description
At the Institute of Robotics and Mechatronics the first telemanipulated robot system used in space was developed for the space robot experiment ROTEX of the D2 mission (STS-55 from 26.04. to 06.05.1993) in cooperation with Dornier (1986-1993). It demonstrated in an impressive manner that it was possible to control a space robotic system from Earth. On-board autonomy was used to compensate for the occurring orbital delays of up to seven seconds. This autonomy was achieved using the multisensory ROTEX gripper. The robotic arm used in the mission was not able to carry its own weight on Earth.
For the training of the astronauts, however, a lightweight and flexible robot was needed: The idea for the first lightweight robot (LBR I) was born. In the first design, the arm had 6 joints and was similar to the flight model of the robot in the ROTEX experiment. Later, a seventh rotational joint was added between axis 2 and 3, providing the robot with a redundant degree of freedom. The kinematics of the robot was similar to that of the human arm, each joint was equipped with a torque sensor, all electronics were integrated into the robot arm, and weight was reduced wherever possible.
The LBR I already had many of the features of the later DLR lightweight robots, which were combined in one robot arm for the first time:
While all methods were pursued further and are now standard in many industrial robots and cobots in different realizations, the real-world goodness of the motion was quite limited. Due to strongly fluctuating friction in the gears, the robot tended to have clearly noticeable vibrations and the running noises were also quite unpleasant. Especially because of this, it was only possible to approximate and slowly display higher dynamics in the overall system, as they occur, for example, with Cartesian force control in the contact case. Nevertheless, it could already be seen with this system how extremely fast joint control in combination with high-resolution sensor technology and high-speed system communication would enable future use cases in the field of assembly and human-robot collaboration. The advantage of kinematic redundancy with significantly more harmonious and efficient motion patterns became visible for the first time as well.
Publications