Institut für
Robotik und Prozessinformatik

Deutsch   English

Zur Zeit leider nur in Englisch vorhanden...

Assisting Robots for a Future Human-Machine-Interaction

Project Description

Demographic changes in the population make it necessary to think about more applications for robots in industrial environments. Tasks like welding or painting in automotive production are already highly automatized. In contrast assembly tasks are mainly done manual by human workers. The average age of these workers is constantly rising. Consequently the ergonomical quality of a workplace becomes more import. One way to improve the ergonomics is supporting these workers with assistive robots. For example, a robot could hold parts in place while a human assembles them.
Due to the shared workspace of human and robot new concepts for human-robot interaction have to be developed. Classic industrial robots are always in seperated cells, in which no human can enter without disabling the manipulator. In a shared environment the robot has to be able to react to his human coworker.

We designed a demonstrator with which we are able to show assisting concepts in scenarios close to reality.

aufbau1.jpg
Fig. 1: Current demonstrator.

We are developing a method which allows a non-expert programming a robot to do a specific task. For this we use a gravity compensated mode in which the endeffector of the manipulator is free floating and can be moved by hand. These movements are then recorded and processed to produce a collision free path for the robot.

assembly.jpg
Fig. 2: Closeup view to complex assembly.

In future we would like to create complex assembly trajectories automatically. In the figure above the cone-shaped bolts have to be inserted into the appropriate hole in the car body, while at the same time the grub screws have to fit into the component. These constraints make it impossible to use a linear assembly trajectory.

input-to-spline.png
Fig. 3: Filtered joint input data and generated spline data.

It took 0.25s to generate this page.