Donald Noble's Website

Google have posted an interesting video of a talk by Sebastian Thrun (who was part of the winning team from Stanford) about their entry into the DARPA grand challenge 2005. See Winning The DARPA Grand Challenge – Google Video – which is almost an hour long.

The DARPA grand challenge was a competition to drive a completely autonomous vehicle across a 132 mile desert course, with no human intervention. More at Wikipedia

It’s interesting to see how the team used a combination of GPS, lasers and video footage to build a 3 dimensional model of the terrain around the car, marking up which areas were drivable, and those that were not (such as rock, ditches and other competitors!). Some of the technical jargon on how they achieved this goes over my head, but looking at the video images of how the car “sees” the world is quite amazing.

Sebastian makes an interesting point at the end of the video, that a highway operating at ‘peak capacity’ has about 90% or more empty space between the cars, because we are such lousy drivers. If computer technology can reduce the spacing between cars, while at the same time improving safety, then that will increase the capacity of the road network without the need to build new roads.

Well worth a view if you are interested in robotics, computers or Artificial Intelligence.

As a footnote, it would be interesting to know how a human driver would have faired in the race—are we obsolete yet?!

This entry was posted on Sunday 19th November 2006 in Computers, Engineering and tagged Computers, Engineering, future, mathematics, technology, transport. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a comment, or trackback from your own site.