Skip to content

emgoeddel/ara-planning

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Implementation of the motion planning agorithm described in:

B. Cohen, S. Chitta, and M. Likhachev, “Single- and dual-arm motion
planning with heuristic search,” The International Journal of Robotics
Research, vol. 33, no. 2, pp. 305–320, 2014.

This is a replication study and was created without referring to code
published by the authors.

This repository contains an implementation of Anytime Repairing A*
search, which can be run over a selection of search graphs. Options
for running the code are below.

2D Any-DOF Arm Simulator
Dependencies: Qt 4, Eigen 3
How to: In top directory, run "make run".
What: Provides an interactive GUI for setting search targets, stopping
and starting searches, switching between search heuristics, and moving
a simulated arm around. Further toggles (ie for changing the DOF of the
arm) are available in the code.

Maze World ARA* Example
Dependencies: Qt 4
How to: In top directory, run "make maze".
What: Demos the functionality of ARA* in a maze world. For more
information about this algorithm, see:
M. Likhachev, G. J. Gordon, and S. Thrun, “ARA*: Anytime A*
with provable bounds on sub-optimality,” in Proceedings of the 2003
conference on advances in neural information processing systems (NIPS),
2003.

Robot Arm Controller
Dependencies: Eigen 3, plus LCM and the RoSIE project, which is not
contained in this repo, to see the search actually control a physical
or 3D simulated arm
How to: In top directory, run "make robot". Accepts and produces LCM
messages defined in lcmtypes directory.
What: A process that communicates with a simulated or physical RoSIE
arm that recieves requests for plans and sends arm commands in return.

Author: Elizabeth Mamantov
2015

About

ARA* search in motion primitive space.

Resources

Stars

Watchers

Forks

Packages

No packages published