Re: darpa show tonite
- From: mimo_545@xxxxxxxxxxx
- Date: 30 Mar 2006 11:01:32 -0800
Curt Welch wrote:
"feedbackdroids" <feedbackdroids@xxxxxxxxx> wrote:
Stanford relied mainly on short-range laser scanning to build
real-time internal representations of the terrain directly ahead and
immediately to the side, and produced a probablistic mapping of good
and bad zones for travel. This was combined with a long-range camera
image of the road several 100s of meters ahead for course prediction.
Then overlaid the internal maps.
I wonder how that worked. Wonder about what happens in the time
between image exposure and mapping. The vehicles/team survey the
course first don't they? - - Don't know if I'm on the right the
track here but - it would seem that it is the combination
of the sensitivity of the sensors to the environment with that
predictive/probability mapping that was so crucial?
Teams didn't know the exact course until 2 hours before the race
started. OTOH, they knew the start and finish line - same place - so
there were only so many roads in the vicinity through the desert they
could travel on. The vehicles were moving up to 40-50 MPH max but more
like 20 MPH nominal, so the maps would have to be computed pretty fast.
Stanford's computer screen map looked like it was being updated at
1/sec. Since the lasers scanners were mapping out to 100+ feet, that
was about enough per screen capture that the vehicle didn't overrun the
front edge of the map before the next update. They'd be in trouble at
100 MPH. Just a matter of having enough onboard computer power [to keep
up].
I haven't seen the show yet (it's on again tonight at 1AM and I've got it
set to record), but last year, some of the groups seemed to have created
their own high-res digital maps of the entire area before hand. They then
spent the two hours before the start hand crafting the exact path and speed
the car would take. The course data as given to them I think defines a
very wide channel they must stay in so by itself, the data given to them
doesn't tell them exactly where the road is. So instead of making the car
smart enough to find it's own way, they plotted the exact course and speed
it would need to take down to a resolution accurate enough to stay on the
road. The sensors then only had to spot problems major problems on the
road and avoid them.
--
Curt Welch http://CurtWelch.Com/
curt@xxxxxxxx http://NewsReader.Com/
Ah Ha!
http://www.pbs.org/wgbh/nova/darpa/program.html
:)
.
- References:
- darpa show tonite
- From: feedbackdroids
- Re: darpa show tonite
- From: mimo_545
- Re: darpa show tonite
- From: feedbackdroids
- Re: darpa show tonite
- From: mimo_545
- Re: darpa show tonite
- From: feedbackdroids
- Re: darpa show tonite
- From: Curt Welch
- darpa show tonite
- Prev by Date: Re: Strong AI Thesis (No Chinese room, I promise)
- Next by Date: Re: What the Hell is a Determinism Anyhow?
- Previous by thread: Re: darpa show tonite
- Next by thread: Re: darpa show tonite
- Index(es):
Relevant Pages
|
|