Robots and Robot Venues: Resources for AI Education
This is a page of working notes for the "one-hour" robot competition/exhibition
planned for the final day of the AAAI Spring 2007
Symposium on robots and robot venues in AI education.
Thoughts on the Schedule
Thoughts on the platforms to use
- Monday, 3/26/07 in the afternoon: Give out the platforms and software during the
- Tuesday, 3/27/07 in the afternoon: Give out three tasks among which particpants
may choose one (or some up with a hybrid or another of their own design)
- Task One: Escape The goal will be that the participant-team's
robot will be placed in an unknown location within a provided map and must escape from
the locations in that map (extra kudos if it knows when it has done so!)
- Task Two: Count The participant-team's robot will be placed in
an environment with a number of distinctly colored objects scattered around it -
the objects will be provided beforehand (probably brightly colored balls). The
robot must count as many balls as it can within the environment (extra kudos here if
it doesn't double-count objects...)
- Task Three: Assign! The participant-team will design and program
a task for their robot that one (or more) participants might imagine using
in one of their classes. Obviously wide open (with extra kudos if the
actually get used!)
- Wednesday, 3/28/07 in the morning: Teams have "one hour" (really more like
two) to work on their task. After this time, each team will present a five-minute
overview of their system to the whole group.
- Wednesday, 3/28/07 at 11am: Symposium wrap-up with reflections on this
"one-hour" exhibition folded into broader thoughts on how robot venues can
most effectively be used for educational ends.
Thoughts on the software to use
- I have enough roombas to supply one for every 3-4 participants, and I can drive
them up from HMC the day before the symposium. Each one can run via bluetooth
(provided with the robots) or with a laptop onboard -- webcams will be provided
for participants to use their own laptops (only Windows/MacOSX - I have no Linux
video knowledge -- Holly?) with vision. Pyro will be the software
interface, though the ascii-string-based socket server will accept connections from
any language with a socket library (i.e., every language).
- Jerry has suggested using XBCs - this would be a great way to get the
ration down to one robot per team-of-two and would allow participants to
choose a platform that would best suit their use at their home institution.
Pyro for the roombas (demoed at FLAIRS 2006), with the extendibility noted above -- I don't know what the XBCs need (perhaps pyro as well?)