Stories
Slash Boxes
Comments

In+ersec+ion for Spatial People

Self-driving Vehicle to Tackle Desert

posted by Satri on Thursday September 29, @12:25PM   Printer-friendly   Email story  Permalink  Trackback URI  Slashdotthis  Diggthis  Del.icio.us
from the one-step-closer-to-the-batmobile dept.
gps writes "A team of Cornell University engineers has built a self-driving vehicle to enter the 2005 DARPA Grand Challenge, in which a vehicle must cross 175 miles of battlefield-like terrain entirely under computer control. The vehicle will be loaded with a set of waypoints which the vehicle must then follow using its onboard GPS unit with no interaction from humans. Sounds challenging? The contestants don't have much faith that anyone will complete the journey."

Related Stories

DARPA's Map-Based Wiki Keeps Platoons Alive [+]
The discussion on Slashdot named DARPA's Map-Based Wiki Keeps Platoons Alive: "One of the biggest problem that a platoon on the ground in Iraq or Afghanistan faces is that when a new unit cycles in, all the street-sense and experience of the old unit is lost. Knowing where insurgents like to plant IEDs, or even which families have a lot of domestic disputes, can spell the difference between living and dying. In response to this, DARPA created TIGR, the Tactical Ground Reporting System. Developed as much on the ground in active warzones as in a lab, TIGR lets platoons access the latest satellite and drone imagery in an easy-to-use map based interface, as well as recording their experiences in the field and accessing the reports of other troops. In this O'Reilly Radar interview, two of the people responsible for the development of TIGR talk about the intel issues that troops face in hostile territory,the challenges of deploying new technology meant for combat areas, the specific tricks that they had to employ to make TIGR work over less-than-robust military networking, and how TIGR is impacting platoons in their day to day operations"
This discussion has been archived. No new comments can be posted.
Display Options Threshold:
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • Cool applications for DEM analysis

    (Score:2, Interesting)
    by spatialguru (105) <tylermitchell@shaw.ca> on Thursday September 29, @01:04PM (#51)
    ( http://spatialguru.com/ )
    I'd like to hear more about these DARPA contestants and how their vehicles make use of computer modelling of the terrain. You'd think that'd be a major part of the mission...
    • Re:Cool applications for DEM analysis

      (Score:2, Informative)
      by Ben (62) on Friday September 30, @03:10AM (#61)
      ( http://vterrain.org/ )
      I've been corresponding with one of the guys doing algorithm development for one of the teams. Surprisingly, standard terrain models aren't too important. For example, for that part of Nevada you could get 10m DEM for the whole area and easily have the vehicle algorithm reference it, but that's pretty useless for actual driving. These vehicles need far more detail - every rock and ditch - which they can only get with realtime capture, i.e. Lidar/Radar. Apparently, the patchy "terrain models" these algorithms have to construct bear little resemblence to anything we would recognize in the CAD/GIS/RS/Vis-sim fields.
      [ Parent ]