What will DARPA think of next?

[mechanicalbrain]

Well as of today Stanford, Red Team, and Red Team Too have reached the finish point in the Mohave Desert! This marks completion of another one of DARPA's ingenious competitions being completed! First the Xrace and now this! So whats up on the horizon I ask? Obviously it has to be something that can be easily (easily meaning with the help of extraordinary sums of money) done. While the coarse was completed I would like to see them do a new one with a time constraint! Also I would like to see a competition involving a personal robotic chassis of some kind. So what do you think DARPA will do next?

[Billfred]

Originally Posted by mechanicalbrain

While the coarse was completed I would like to see them do a new one with a time constraint! Also I would like to see a competition involving a personal robotic chassis of some kind. So what do you think DARPA will do next?

There was a time constraint--ten hours.

Maybe I'm screwed up in the head, but I think the next level will involve dealing with more traffic. After all, if the military's looking to put autonomous vehicles in combat, they're going to have to deal with things moving around them.

Could it be that they'll combine the two? unmanned vehicles going to near-space?

Or, due to the fact that it is for military purposes, maybe automated weapons of some sort... geez that would be dangerous.....

[mechanicalbrain]

Yeah I don't see weapons happening but they will definitely keep to their pattern and do something involving military research like Xrace and the Grand Challenge!

[EricH]

How about this one: something similar to the Grand Challenge, but at certain points, you have to stop and put x number of projectiles (say, rifle or machine gun bullets) into a target. Said target may be anywhere off or on the road, and a human is in the area to replace it or move it. Reloads must be autonomous and the vehicle must have said weapon inside, but not break a window.

Oh, and the X Prize was not done by DARPA. If I remember correctly, it was done by the X Foundation, or someone with a similar name.

Now that I've done some thinking on it, the weapons being involved is not the smartest idea... imagine how much damage a single software glitch would cause!

[Matt Leese]

Originally Posted by Billfred

Maybe I'm screwed up in the head, but I think the next level will involve dealing with more traffic. After all, if the military's looking to put autonomous vehicles in combat, they're going to have to deal with things moving around them.

They're not looking for combat vehicles. They're looking to make automated resupply vehicles. This is particularly pertinent given that supply convoys were being heavily targetted by the insurgency in Iraq. If you don't have actual people driving the trucks, it becomes a lot safer to run the convoys. It's also much cheaper for the military (people are expensive). Remember that for every one combat soldier there are something like 10 soldiers behind the lines to make sure that one can fight (those numbers stem from World War 2; I'm not sure what the current figures are but I'd bet them to be higher if anything).

Matt

[Billfred]

Originally Posted by Matt Leese

They're not looking for combat vehicles. They're looking to make automated resupply vehicles. This is particularly pertinent given that supply convoys were being heavily targetted by the insurgency in Iraq. If you don't have actual people driving the trucks, it becomes a lot safer to run the convoys. It's also much cheaper for the military (people are expensive). Remember that for every one combat soldier there are something like 10 soldiers behind the lines to make sure that one can fight (those numbers stem from World War 2; I'm not sure what the current figures are but I'd bet them to be higher if anything).

Matt

Granted, bad word choice. But even if all war were to end tomorrow, and we could all frolic in the fields, such devices would still have to deal with traffic.

[Mike]

Originally Posted by mechanicalbrain

Are you sure? It looks like some teams are going well into 11 hours.

Well I'm sure that the 10 hours is to win the competition. They will probably let robots run until they complete the course just to see what time they will get.

[dlavery]

The conditions of the Grand Challenge require that the robots finish the course in 10 hours or less to be eligible for the $2,000,000 prize. Operations are continuing on October 9 to allow TerraMax to complete the course, even though they will be out of contention for the money (given that not even one team finished last year, just being able to finish this year is a very worthy achievement, so they are being given every chance to do so by DARPA).

-dave

[Travis Hoffman]

Originally Posted by 114ManualLabor

Could it be that they'll combine the two? unmanned vehicles going to near-space?

Or, due to the fact that it is for military purposes, maybe automated weapons of some sort... geez that would be dangerous.....

Cue Terminator 2 theme...

We don't need THAT kind of science fiction to become reality, thank you very much!

[Rich Kressly]

The "status board" page at http://www.grandchallenge.org/ has the most up-to-date and complete info thus far.

Currently TerraMax is on the move with 106 miles complete.
Four teams completed the 132mi course yesterday (times may still change):
Stanford - 9h 55m
Red Team - 9h 59m
Red Team Too - 10h 4m
Gray Team - 10h 17m

As Dave said, DARPA is giving TerraMax a chance to finish, a team has to be under 10h to be considered for the 2M, and we should know later today (after all times are finalized) who the winner is. Congrats to all of these teams. The improvements in only a year are amazing.

[Mike]

Originally Posted by 114ManualLabor

Could it be that they'll combine the two? unmanned vehicles going to near-space?

I doubt that will happen anytime soon. Think of all the land vehicles that got software glitches and went off-course during DARPA. Now when they are flying, instead of going off-course they fly into communities and cities nearby.

[Matt Leese]

Originally Posted by Billfred

Granted, bad word choice. But even if all war were to end tomorrow, and we could all frolic in the fields, such devices would still have to deal with traffic.

Not really. They're not designed for an urban environment. More for resupplying troops out in the field. There isn't too much traffic when travelling offroad.

Matt

[dlavery]

Originally Posted by Rich Kressly

Currently TerraMax is on the move with 106 miles complete.
Four teams completed the 132mi course yesterday (times may still change):
Stanford - 9h 55m
Red Team - 9h 59m
Red Team Too - 10h 4m
Gray Team - 10h 17m.

Note that the times shown on the web site status board are the elapsed track times since the robots started on the course, and not reflective of the actual time to complete the course. The clocks keep running after the robot passes through the finish chutes.

I was in Primm, NV, for the start of the race yesterday, and got to see the first few vehicles start out onto the track. CMU's Sandstorm had the fastest run time during the NQE, but they let H1ghlander go first, then Stanford's Stanley, then Sandstorm. Stanley passed H1ghlander around mile 109, which meant that it had at least made up the 5-minute difference in the starting times, and was running faster at that point. But, since Sandstorm started after Stanley and they both came in with about the same separation in finish times, it was not clear which one actually ran to course fastest. We are still waiting for a posting of the official completion times from DARPA.

It is worth noting that there were sections of the course where the vehicles had to contend with 2-way traffic. The vehicles were entering the course at staggered starting times. There were locations on the course where the "out and back loops" utilized the same sections of road. So there were many occasions where the early starters had to contend with oncoming traffic as they met their later starting competitiors on the course.

-dave