Graphics enhanced version of this site
Navigation Among Movable Obstacles (NAMO)
Heads: Chris Atkeson and James Kuffner
Contact: Michael Stilman
Mailing address:
Carnegie Mellon University
Robotics Institute
5000 Forbes Avenue
Pittsburgh, PA 15213
Associated center: CFR
Associated lab/group: Planning and Autonomy Lab
For more information, see this project's homepage.
We are developing algorithms and methods to allow humanoids and other dexterous mobile robots to interact with their environments. In particular we are interested in two aspects of this problem:
How can robots reason in real-time about environments consisting of numerous rigid objects? Given a navigation or manipulation task, a robot may have to manipulate a number of unspecified objects. For instance, objects may lie in the path of the robot and will need to be removed. Furthermore, some objects could be useful and should be exploited by the robot to accomplish the task. We approach this problem from the perspective of motion planning, and explore low dimensional subspaces that contain the most relevant information about operators to conform the environment to the robot's goals.
In order to perform safe, repeatable and robust environment interaction we are also interested in developing automated methods for robot control. We seek to construct controllers that maintain safety guarantees in environments that require complex contact, underactuated dynamic motion and limited sensing. Again we are interested in reducing the dimensionality of the problem towards feasible computational solutions.
Note: This list may not be comprehensive. It contains only those publications in the RI publications database. Entries are listed in reverse chronological order.