Lockheed Martin wants a fingertip-sized UAV in every soldier’s backpack | Hacker News Reader

Lockheed Martin wants a fingertip-sized UAV in every soldier’s backpack

geek.com

40 pointsukdm26 comments

Open articleView on HN

Interesting fact from the second video: The vehicle is controlled by a Honeycomb app.

"Lockheed Martin wants billions of dollars"

Well, no shit.

Did they change the article? I didn't see that in the text or videos.

I was just rephrasing the headline.

Poorly.

Billions seems unlikely. 450k enlisted men, 100k total force in Afghanistan. Let's guess the Army will want 250,000 of them. You'd need to sell them for $4k each to get that first billion dollars.

Sure, you can sell the first few for more than that during testing, you have to recover the cost of the documentation and supporting materials the army will require plus pay all the people you hire to interface with their counterparts, but ultimately I think you'll end up in the $1k neighborhood and still be making a huge per unit profit. ($200 would be a nice price for the consumer version. Maybe more depending on how nice a camera you put on it.)

(I assume the soldier will already be carrying a suitable display.)

Hmm fair enough.

Don't forget police forces. $4k is not out of reach for a medium sized metropolitan police force...and there are many many many of those.

Now that the video is out, I'm sure some production lines are coming online in China as we speak for a $20 knock-off, maybe $50 for one with a camera.

Meanwhile, the incredibly nimble Syma S107G helicopter sells for $20 shipped on Amazon (affiliate).

http://www.amazon.com/gp/product/B004A8ZRB0/ref=as_li_qf_sp_...

If you add the infrastructure needed to live combine the video feeds of say 1000 units into a coherent view, available to each of those units and their commanders, but not to the enemy, things can get expensive.

Having said that the combination of 'finger-sized' and outdoor operations will be quite a challenge. Even moderate wind will be a problem.

What's the advantage of using the asymmetrical one-wing design which undoubtedly makes it harder to control?

"Notice how stable it hovers. This exploits the natural stability of maple seeds" (erm.. I tried to quote as good as possible).

And it seems to reduce the complexity of the device itself a lot. One motor. One wing. I looked at it and thought "Wow! What a great and 'simple' (as in reduced) idea".

It's basically the simplest version of a helicopter possible where the entire craft is the rotor.

It's pretty much impossible fly fixed wing UAVs indoors. This concept is pretty amazing.

I read that too. I just don't believe it. I think a two wing version with the same dihedral angle would be just as stable. Their design might eliminate the "complexity" of one motor though maybe they can do two wings and one motor. Unclear whether the difficulty in controlling it (which they also mention) is worth the tradeoff. There is the gee whiz factor though.

I think the complexity in controlling it is due to the fact that the whole craft is rotating (where's the front..?). If you want to remove that complexity, you need to add a lot more than just another wing and add a number of mechanical parts that can fail. As it is right now, the whole complexity is the software (which should be safely stored).

In addition, this design lends itself to small sizes in a great way. Shrink the center parts/electronics/, shrink the single wing and you get a very, very compact tool. Another wing nearly doubles the length of the whole thing.

The other wing would fold next to the first wing. Since there would be more wing area, the first wing could be made smaller. Not necessarily any additional moving parts needed.

which undoubtedly makes it harder to control

That's a problem for the software. Divided over the number of proposed units, the development cost is about zero. The mass of the zeros and ones is also nearly zero.

What it saves is moving parts, actuators, and driver circuits, all of which have a mass and cost per unit.

!!! I make a guess on how it operates here, I think it's good, but don't take it as gospel !!!

I didn't see it spelled out, but I theorize what they do is modulate the thrust of the single motor by the wing's rotational position to tip the entire object causing it to trade some vertical thrust into lateral thrust. Given that they already had to have a micro controller and a variable speed motor driver for the motor, the incremental cost in mass and dollars is zero.

Smaller: NAV (Nano Air Vehicle) by Aerovironment - engineering folks behind the General Motors EV1

bottom-right corner of their homepage: http://www.avinc.com/

They didn't make it clear in the interview if they'd solved the imaging problem or not yet. How do you get usable video (or stills) back from a camera that's rotating constantly - and potentially, erratically.

Perhaps they could have the camera counter spin.

Can gyro / accelerometer data be used to give the display client hints on what compensation to make?

He didn't explain how and to what extent, but in the second video he does say that they managed to "solve" that problem.

Ah so he does - although no video from the 'copter.

In the second video, they said they'd solved it, though not how, nor was any imagery shown.

Counter-rotating the image, synced to the rotational speed of the device, would seem to be a simple solution, provided the camera was centered i.e. it wasn't constantly panning.

Even if it was (e.g. mounted on the fastest-moving outer edge of the wing), software could reconstruct a panoramic view. You don't even need to align images from subsequent rotations, if you know what direction it is pointed. You'd get fps related to RPM though.

When flying it looks eerily like a classic flying saucer. It even has a sci-fi-ish tone to its sound.

Reply on news.ycombinator.com