DARPA

[falconmaster]

You got it man! Free burgers on the house!

[Adam Shapiro]

I thought I saw some 842 shirts in the crowd, but we were all too nervous for our car to think of anything else. I hope you guys enjoyed the event as much as I did, it certainly was an amazing experience.

[JBotAlan]

I haven't been tracking this closely enough to know, but I do have a question: how hard would it be to start a DARPA team at a school that doesn't have one? What is involved? Obviously a corporate sponsor; what else?

JBot

[Chief Pride]

i have been looking for a video of what happened since the 3rd, i still cant find anything... and i can usually find ANYTHING! :s

[Adam Shapiro]

Originally Posted by JBotAlan

I haven't been tracking this closely enough to know, but I do have a question: how hard would it be to start a DARPA team at a school that doesn't have one? What is involved? Obviously a corporate sponsor; what else?

JBot

There have been a number of threads in the past about just this. In fact, there is a recent thread in which a number of people have expressed interest in creating an entry by joining skills from a number of FIRST teams.

Personally, I would say that it is a very difficult task for anyone, and would be rather difficult for a FIRST team to undertake. I'm certainly not saying that it couldn't be done, and if you are really interested in it then by all means go ahead, however from my experience it would be near impossible to find enough hours in the day to work while still in high school (not to mention still having to work on the robot!). Our team consisted of about 13 members, each of which worked at least 20 hours a week on the project. In the last 6 months before the event most of the team worked on the car full time, putting in about 12-16 hours a day of straight work to get the project completed. A few team members even had to take this semester off to finish the AI and prepare for competition.

The complexity of the software required to control the car also makes it a very hard task. Actuating the car and running the wiring is a minor task compared to the work that goes into making the car think and drive on its own. Our own pose estimator (position and orientation) took about a year and a half to get fully functional, and required processing and fusing two different GPS units, a missile-grade IMU, and local world data from a number of sensors on the car.

That being said, the DARPA competition is an amazing event and if you're still interested in entering (if there is another one) I say go for it. I guarantee, if nothing else, you'll learn a ton and have a great time doing it.

[Molten]

I think the fact that it is not on the news, says more about the viewers then the media. Afterall, the media shows what "people" want to see. I just find it sad that people would rather see who got drunk at a hollywood party then the future of the automotive technology. (Oh well, I will always be to much of a tech head to understand society.)

[kE7JLM]

Originally Posted by JBotAlan

I haven't been tracking this closely enough to know, but I do have a question: how hard would it be to start a DARPA team at a school that doesn't have one? What is involved? Obviously a corporate sponsor; what else?

JBot

Well, Stanford's cost $500,000. It took a year to build. Also the had to identical cars.

[Adam Shapiro]

Originally Posted by kE7JLM

Well, Stanford's cost $500,000. It took a year to build. Also the had to identical cars.

Don't forget that Stanford also had considerable help from VW, who both provided vehicles and implemented a full drive-by-wire system for them. A good deal of money is also needed for overhead for the project. It doesn't all go directly into the car. We received $1M from DARPA (as did Stanford) as one of the Track A teams, as well as some additional funding and assistance from other sponsors. A large part of that money did not actually go into the car in the end. I'd say our car cost somewhere around $600,000-$700,000 in the end, but it's hard to tell.

As for the length of time it took to build, most teams (not all though) have been working on the project since the competition was announced last April. There is much work that has to go into the project long before work on the car itself ever begins. We began planning towards the end of last April, and actually purchased our car in January. Actuation and control took about two months to develop (though revisions continued until about July), sensor interfacing took from last August to around June, and AI was under development from before the competition announcement until the day the car left for CA. The total number of man-hours that went into this project is no longer a number we can calculate, but when people say the project took a year and a half of their life they mean it.