Last night's session was the most productive to date. I found a thread on Darpa's Board that answers many of our questions about why there may have not been a winner for this contest. I recommend *everyone* to read this thread..this is completely new information that is nowhere else. In fact, I made a copy of the entire thread in case it disappears. This is really very enlightening stuff!

darpa.mil/grandchallenge

click on discussion boards>click all the agreements>

Go to Forums>Position Lcation-GPS

The thread is called "Why did everyone fail?"

OH...paydirt. Ain't the web great?

http://www.intelligententerprise.com/000120/decision.jhtml

Hi! Geez. Des is becoming a regular thing with me...this all- nighers are very productive but...this ain't no way to make a livin. Anyways...re-read William "Red" Whittaker's racelog thing...and I've got to admit something. These people are good...*really really* good. Mr. Whittaker even found some humor about the wreck that happened 10 days before the race that I believe cost them the race...or at least contributed heavily to what happened. I'd still like to know more about what happened there on Daggett Ridge...more details..what was Sandstorm "thinking"? I've been putting a tremendous amount of time coming up the "best" design for everything...but the primary CPU is my greatest concern. I'm not sure how much has been written about this...but...it seems to me if we are going to design machines that make decisions, we'd better have a way of back-tracking those decisions build in..like a log. "At 10:23 I made the decision to increase speed 10%" will not be enough by itself without a "why" along with it. It is clear from what I've read they did what they could after the wreck to fix everything...but it just was not possible. Why? Sadly, I firmly believe they never thought this could or would happen. I'm sure they know what happened..but..they may never tell us. $3Million dollars worth of hardware and software...incredible. Maybe we'll get lucky and after they know the truth...they'll share it with us. Anyways..truly one of the most advanced robots built to date.

Pulled another all-nighter researching and thinking about the problems I'll encounter with this contest project. Sadly, I found out how much some of these gadgets cost! Starfire is *not* cheap...about $10K with a year subscript. Gee..I sure hope John Deere likes sponsoring people! :) Rethought the logistics of having two vechicle entries...prolly better off with one fully-tested, well- thought out vechicle. I found out those cool John Deere Gator vechicle's max speed is about 25MPH...perhaps a special engine/trany? I'd really like to have more speed than that..just to be on the safe side. I've started designing prototype1...a small basic robo vechicle to jump start software design and see what I'm up against with all this. I think good ultrasonics are going to be key. I discovered a company called Sonin...these little distance measuring critters rock. They're cheap...~$50...and unlike many of the other ultras, they have a range of 60 feet. Not sure how easy it'll be to interface with the microcontrollers..but time will tell. I can't wait to get one of these. Oh..yay! My first cert! The system still won't let me pst the article from yesterday...perhaps two people have to give me a cert. Anyways...more laters. I'm pulling another all-nighter and that will be all for awhile..quite literally. Decided to see if I could get a laugh out of Darpa by submitting a pre-app team name request reservation.

As I said in the rant, the key to winning the race will not be who has more or better toys or what-not. No. The answer is this: Less everything. Fewer bells and whistles both in hardware and software...and better, simpler software that does the job it needs to do..and do that job well...day after day...week after week...year after year...everytime you turn it on. Not only good sensors..but a well-thought out system to protect them when need be and constantly test them. I think a completely redundant main CPU would be a very good thing...one that could be designed with a fall-over feature.