Adaptive Walking Pattern Generation with Damaged Legs for the six-legged Robot LAURON V

C. Plasberg,D. Kogel,A. Roennau,R. Dillmann
DOI: https://doi.org/10.1109/icar53236.2021.9659374
2021-12-06
Abstract:To get to remote locations in challenging environments, autonomous robots have proven to be a great tool. Walking robots are one of the most exciting robot types suitable to tackle the challenge of reaching those locations, crossing all the encountering terrains, and even climbing over obstacles. As the locations get more and more remote or even are located on other planets, repairs get costly or impossible. To maintain mobility in case of damage, strategies need to be developed to handle upcoming defects. Hexapods offer the advantage of redundancy in their locomotion apparatus. This redundancy can be used to deal with defects in its legs. The here presented approach deals with defects in one leg without detaching the damaged leg. At the same time, a statically stable pose is guaranteed at all times. Superior knowledge of the damage or the position of a damaged joint is not required. Two different walking procedures are developed that enable the robot to walk without the damaged leg, one to increase speed while the other focuses on stability. The concept is implemented and tested on the six-legged walking robot LAURON V.
What problem does this paper attempt to address?