Project Criteria
Here are the guidelines I mentioned
in class on Monday Feb 22:
-
Projects should focus on some topic related to manipulation or mobile
robotics and are intended to be an effort over approximately 5 weeks
-
I encourage teams of 3 +/- 1
-
There will be some project update meetings on several mondays as
per the syllabus
-
Written report and class presentation are due at the end of the
semester
-
Tentative project proposals due Friday March 5
-
Scope and suitability of projects will be negotiated between me
and project teams
-
Projects need not be ground-breaking research.
More moremore
-
Some students have been coming to me with ideas for projects that have
been focused on building neat robots rather than on interesting "scientific"
problems in robotics, so I suggest the following guideline:
There should be more science than engineering in
your project.
-
The focus of your project should be on some "scientific" problem, which
is not to say that there can't be any engineering, but that it should play
a secondary role, e.g. construction of hardware to implement or verify
a solution to a "scientific" problem. If you were pressed for time
at the end of the semester, it should be the engineering that gets left
out rather than the science. The engineering should not be critical
to your project. "Engineering" can apply to software as well as hardware.
By "scientific" problem, I mean a problem in which has some generality
and for which you will take a principled approach in a solution.
-
One of the questions that I will be asking as I evaluate the suitability
and scope of projects is what you will learn from doing this project.
-
I will tend to expect a little more from students who are registered for
the graduate version of the class. Let me know if there are students
on one project team who are registered for different versions of the class
(some for 4964 and some for 6963).
-
You might also look through the International Conference on Robotics and
Automation, IEEE Transactions on Robotics and Automation, and the International
Journal on Robotics Research for ideas.
Project Suggestions
Hardware projects
-
Aside from the nomad robot, i expect to have computer platforms that have
A/D, D/A, and DIO capabilities, some motors, and some solenoids.
I also expect to have a camera and framegrabber. Hobby servos are
a possibility. Other hardware is a possibility.
Software projects
-
You can write a simulator for almost any sort of robotic system and then
experiment with algorithms to control that system.
-
There is a robot soccer simulator available, but this may be an ambitious
project...
-
Implement some algorithm from a paper. Some possibilites are:
-
A nonholonomic path planning algorithm
-
Planners for modular robots (such as those by Rus, Chirikjian, and
Unsal)
-
Parts feeding systems - design a scheme for feeding parts and develop
a simulator and algorithms
-
Modular fixturing algorithms
-
The Coulomb pump (using time averaged frictional forces to position
parts on a surface)
-
I have a simulation environment for multiple mobile robots available if
you are interested in collaborative robotics
Term paper projects
-
Some aspect of mobile robot navigation or localization
-
Various control topics related to robotics (mostly concerning underactuated
systems)
-
Visual servoing
-
Cooperative robotics