Evaluation of human-likeness of robot arm motion trajectories.

In robot arm applications involving human-robot interaction, ensuring safe and predictable robot motions is of critical importance. These qualities can be enhanced by designing motions that resemble human movements. Previous work at ASM has focused on creating trajectories derived from human motion data and analyzing how human muscles produce movements. This work should build upon these existing methods to further improve the generation of robot trajectories. For both approaches, scenarios are to be defined where the robot arm interacts with objects or the environment. Experiments should then be conducted to implement these methods and compare their performance. Furthermore, a study involving human observers should be carried out, using questionnaires to evaluate the human-likeness of the robot’s motions.

More information can be found here.

Kategorie: Master Thesis