Stanford Code From Cars That Entered DARPA Grand Challenges
sourceforge.netThis code is legit (see other comment; download the tar and try to forget you ever visited Sourceforce) and incredibly interesting. I browsed this repo a year ago and still come back to it every so often. The architecture decisions that the Stanford team made in 2005 and 2007 have stood the test of time (at least as evidenced by many of the decisions in the architecture of our Voyage car, but anecdotally in others too).
Learn more with this paper on Junior, which uses this codebase: http://robots.stanford.edu/papers/junior08.pdf
Hi Oliver, I have enrolled for the Udacity self-driving car Engineer Nanodegree and cannot wait to start! Do you think that browsing through this codebase will help me better understand some of the concepts taught on the course?
One hundred million percent yes.
Most of this content isn't covered until Terms 2 & 3, but it's surprisingly relevant.
> Stanford Made Available Code From Cars That Entered Darpa Challenges
That's oddly worded. The wording makes it sound like it's the code from numerous different cars in the challenge, and not just Stanford's entry, but the site wording and the source structure itself look to be a single system.
Then again, the code is from 2011, so I guess the past tense is somewhat justified...
Ok, we've replaced the title above with a grammatical one that we ripped off of https://news.ycombinator.com/item?id=14651524.
my bad, ESL, thanks for your "grammatical" rewording though.
Nah no worries, we're delighted that HN has users from many nations. And it wasn't my rewrite, I stole it :)
This looks like an updated version of the DARPA Urban Challenge code. It uses ROS extensively, and ROS was only started in 2007. The files are all dated 2011.
Not for use at high altitudes. From "perception/trafficlights/src/traffic_light_view.cpp:
sprintf(lightBuffer,"%f %f %f", currPose.latitude, currPose.longitude, currPose.altitude);
The perception end of things is mostly digesting the point cloud from the Velodyne LIDAR. There's vision code for traffic light recognition, but I haven't found other vision code yet.
Why can't this be used at high altitudes?
Perhaps laser behaves differently when there is less atmosphere between it and the target.
Maybe sprintf("%f") doesn't give you enough digits, or something along those lines.
Since when does anyone post things on sourceforge anymore? It makes me wonder how legitimate this is...
In Jan 2009 when it was created? Plenty of people. Github was about 8 months old at the time.
It was posted many years ago
I can agree with that. Even if the usernames (and real names) match members from the team, it is still a little suspicious
I can verify it's legit (Sebastian Thrun sent this link to me a year or so ago).
Interesting that they used ROS. I am surprised it was baked enough back then to be useable on a self driving car project.
Patiently waiting for Animats to tell an interesting story about this code :)