Advanced Mobile Manipulation Challenge
Mobile Manipulators are robotic systems that are capable of
navigating through and interacting with unstructured environments in
order to perform a wide range of tasks. Designing control systems for
this domain is a non-trivial process, due in part to the complexities
of constructing and manipulating flexible and extensible
representations of the environment, motor action and task. In this
seminar course, we will study the representation, learning,
planning and control techniques that are currently driving this field
forward. This work will be done in the context of the
DARPA Robotics Challenge.
When: Tu/Th 1:30-2:45
Where: Felgar Hall 300
Topics will include:
- Integrated planning and execution of whole body movements.
- Sensor processing and model construction.
- Planning under uncertainty and with incomplete information.
- Grasp/manipulation planning and execution.
- Human-robot interaction for mission-level control.
This class is a project-driven class in which students and faculty
will work together to design and implement a solution to the
Robotics Challenge. In particular, our goal is to have a
competitive solution in time for the Virtual Robotics Challenge
qualification rounds that will be held in early May.
The students and faculty will collaborate in setting project goals and
milestones, and in determining the content of lectures and class
Students will be required to:
- Read self-selected papers that are relevant to solving the DRC problem.
- Engage in our common semester-long project.
- Both lead and participate in technical discussions, and
report project results in both written and oral form.
Permission of one of the instructors is required to enroll in this course.
Students need to have a solid background in robotics and the
Robot Operating System.
fagg - cs ou edu
Last modified: Mon Jan 14 02:18:26 2013