A Practical Comparison of Motion Planning Techniques for Robotic Legs in Environments with Obstacles

ATHLETE is a large six-legged tele-operated robot. Each foot is a wheel; travel can be achieved by walking, rolling, or some combination of the two. Operators control ATHLETE by selecting parameterized commands from a command dictionary. While rolling can be done efficiently, any motion involving st...

Full description

Saved in:
Bibliographic Details
Published in2009 Third IEEE International Conference on Space Mission Challenges for Information Technology pp. 155 - 162
Main Authors Smith, T.B., Chavez-Clemente, D.
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.07.2009
Subjects
comparison
Dijkstra
Foot
hexapod
Hip
Leg
legged
Legged locomotion
lunar robot
Mobile robots
Moon
Motion planning
Orbital robotics
probabilistic planner
Robot kinematics
SBL
walking
Wheels
Online AccessGet full text

Cover

More Information
Summary:ATHLETE is a large six-legged tele-operated robot. Each foot is a wheel; travel can be achieved by walking, rolling, or some combination of the two. Operators control ATHLETE by selecting parameterized commands from a command dictionary. While rolling can be done efficiently, any motion involving steps is cumbersome - each step can require multiple commands and take many minutes to complete. In this paper, we consider four different algorithms that generate a sequence of commands to take a step. We consider a baseline heuristic, a randomized motion planning algorithm, and two variants of A* search. Results for a variety of terrains are presented, and we discuss the quantitative and qualitative tradeoffs between the approaches.
ISBN:9780769536378
0769536379
DOI:10.1109/SMC-IT.2009.26