As the DARPA Grand
Challenge approaches, articles about teams and their robots are
begining to show up in lots of local papers. The Baton Rouge Advocate
has an article
on University of Louisiana at Lafayette's CajunBot.
CajunBot also rated an article
in the Lafayette Advertiser. You can read about the University of
Florida's Team CIMAR
Navigator in a recent newswise
article. The Pittsburgh
Post-Gazette covers the CMU Red
Team's Sandstorm. And the San
Francisco Chronicle recently covered the three teams from the Bay
Area: Team Digital Auto
Drive, Team Overbot, and the Blue
Team.
I love the red team. I work right down the hall from Red‛TM]s office.
I couldn't join them when I started at CMU last June because I was
already committed to an advisor, but their work is excellent.
A few thoughts:
-Using the most advanced perception, combined with extremely detailed
maps and pre-planning is the most likely to succeed.
-The stabilization for the computation is amazing: the vehicle itself,
a custom stabilization mount, internally stabilized racks, disks in
suspension, and robust disks for internal stabilization.
-Their plans to continue work after the race, regardless of the
outcome, is respectable and exciting.
-Their win would greatly influence my chances for a good job :)
_ivan
Red team also is tackling the problem from a bigger picture perspective
(mapping, route planning)instead of just a react and navigation
perspective (which they also are doing), but then they have the
resources to do so. Which hopefully would be the case also in a live
military action which I believe is DARPAs main usage scenario.