Hummingbird Ornithopter UAV. Yes, you read that right.

[weemadando]

First the AI Penguin Blimps and Submersibles from Festo, and now DARPA man up and bring us bird-sized ornithopter UAVs. When did the world suddenly decide to go nuts with making the crazy tech that we were promised as children?

Video: Pentagon’s Robo-Hummingbird Flies Like the Real Thing

* By Noah Shachtman Email Author
* July 2, 2009 |
* 1:33 pm |
* Categories: Bizarro, DarpaWatch, Drones



Military-backed researchers have built a tiny drone that looks and flies like a hummingbird, flapping its little robotic wings to stay in the air. So far, the mock bird, built for Pentagon mad-science division Darpa, has only stayed aloft for 20 seconds at a time. But that short flight was enough to show the potential of a whole new class of miniature spies, inspired by nature. Darpa just handed AeroVironment, makers of the winged “nano air vehicle,” another $2.1 million to build a hummingbot 2.0.

Ultimately, Darpa program manager Todd Hylton says in a statement, he’d like see “an approximately 10-gram aircraft that can hover for extended periods, can fly at forward speeds up to 10 meters per second, can withstand 2.5-meter-per-second wind gusts.” He also wants the nano air vehicle to operate inside buildings, and be controllable from up to a kilometer away.

AeroVironment, for its part, doesn’t just want its little drone to fly like a bird. The company wants the thing to look like one, too:

[Memnon]

In addition to being used for surveillance, the drone looking this much like a bird could help us study bird behavior in some meaningful ways (like hummingbird communication). I wonder how scalable the technology is - after all, once you get past the hummingbird, wings become multiple-jointed.

[Broomstick]

Oh, it's scalable alright - we own a couple RC ornithopters right now. Ours are more toys than sophisticated spy drones. The technology is maturing rapidly and becoming less expensive. It's not so much the airframe that's improved but the battery/power storage technology. That was the limiting factor.

[Sea Skimmer]

We don’t necessarily want sophisticated out of a small drone like this, such things are going to break so often hitting things indoors that its best to make it as cheap as possible and expend a new one almost every single mission. The US Navy is testing a 100% expendable miniature UAV as we speak, they determined that it was actually cheaper to expend a drone (launched out a sonar buoy tube) then to have a maritime patrol plane burn the fuel necessarily to fly down low to inspect ships and then climb back to cruising altitude. Similar systems are desired to be launched off vehicles, fried from artillery and simply thrown into action by troops. In the future many UAVs will just be a kind of ammunition, to be expended with as much thought as troops might give to firing an anti tank rocket.

[Sky Captain]

This is really cool. I`m wondering what kind of flight time and distance are theoretically possible for such small UAV`s. Battery life are going to be a limiting factor, but what could be possible with the best possible battery technology?

[Akkleptos]

This is really cool. I`m wondering what kind of flight time and distance are theoretically possible for such small UAV`s. Battery life are going to be a limiting factor, but what could be possible with the best possible battery technology?

Exactly. Sorry to burst any bubbles, but at 20 seconds fly time it's not going to get very far. Okay, so they might make a model that flies... uh... 5 minutes on battery power... Just how far is that 20 cm plastic bird going to fly in that time?

Nevertheless, the technology does show promise.

[Broomstick]

We can get 10-15 minutes out of some of our small, battery-powered RC's - but they're just flying machines and aren't required to have a camera or any sort of other tech on board which helps considerably with the weight and power requirements. They're also bigger than this hummingbot, which allows for a larger battery.

I don't know if the military has access to batteries smaller and, for these applications, superior to what the civilian market has.

[Beowulf]

Well, one boost to power that they can manage is to use primary batteries instead of rechargables. They should be small enough that replacing the battery isn't going to be much of a logistics burden, and primary batteries typically have much greater power capacity compared to rechargeables. Example: Lithium-iron batteries have around 3-4 Wh of power capacity in the AA size at the rather high discharge ratings we're probably looking at. Alkaline batteries have more like .8 Wh for the same size.

[Memnon]

Well, one boost to power that they can manage is to use primary batteries instead of rechargables. They should be small enough that replacing the battery isn't going to be much of a logistics burden, and primary batteries typically have much greater power capacity compared to rechargeables. Example: Lithium-iron batteries have around 3-4 Wh of power capacity in the AA size at the rather high discharge ratings we're probably looking at. Alkaline batteries have more like .8 Wh for the same size.

Hang on, what?
If you're talking about lithium-ion, it's rechargeable. The Lithium-Iron-Phosphate chemistry (LiFePO4) actually has more recharge cycles than Lithium-Cobalt Oxide (the usual, LiCoO), but less power. Alkaline is generally not recharged.

EDIT: I assume you mean LiFePO4 by 'lithium-iron.'

[Beowulf]

Well, one boost to power that they can manage is to use primary batteries instead of rechargables. They should be small enough that replacing the battery isn't going to be much of a logistics burden, and primary batteries typically have much greater power capacity compared to rechargeables. Example: Lithium-iron batteries have around 3-4 Wh of power capacity in the AA size at the rather high discharge ratings we're probably looking at. Alkaline batteries have more like .8 Wh for the same size.

Hang on, what?
If you're talking about lithium-ion, it's rechargeable. The Lithium-Iron-Phosphate chemistry (LiFePO4) actually has more recharge cycles than Lithium-Cobalt Oxide (the usual, LiCoO), but less power. Alkaline is generally not recharged.

EDIT: I assume you mean LiFePO4 by 'lithium-iron.'

No, I meant the LiFeS2 chemistry used in lithium disposables, like you can get from Energizer. You might be able to get a AA size battery that's lithium ion, but it wouldn't work in anything that took AAs.

[Broomstick]

No, I meant the LiFeS2 chemistry used in lithium disposables, like you can get from Energizer. You might be able to get a AA size battery that's lithium ion, but it wouldn't work in anything that took AAs.

Unless you're my Other Half, who would alter the device to take the lithium ion. In fact, he's done that. There was also the time he jump-started the car with a mobile phone. If my resident mad scientist can pull that off I presume DARPA and the US military can do so as well.

[Beowulf]

No, I meant the LiFeS2 chemistry used in lithium disposables, like you can get from Energizer. You might be able to get a AA size battery that's lithium ion, but it wouldn't work in anything that took AAs.

Unless you're my Other Half, who would alter the device to take the lithium ion. In fact, he's done that. There was also the time he jump-started the car with a mobile phone. If my resident mad scientist can pull that off I presume DARPA and the US military can do so as well.

Well, it's most likely they'd just choose a battery, and design the power circuitry around that. Anyways, the point is that primary batteries still have better power density than rechargables. Rechargables are more popular for a lot of circumstances, because few people want to go buying new batteries for their phones every other day, for example.

[Shroom Man 777]

Maybe if these small birds become easily mass-reproducible, replacing the batteries would be a non-issue since the UAVs themselves would become disposable?

[Broomstick]

Not to mention that if the military starts mass-producing them then engineering and manufacturing batteries designed specifically for them makes a lot more practical sense.

[TheLostVikings]

Maybe if these small birds become easily mass-reproducible, replacing the batteries would be a non-issue since the UAVs themselves would become disposable?

That was my impression of where we're headed as well. You'd pull the pin on the disposable UAV "grenade" and toss it into the suspicious building/room while watching on a (nondisposable) reviver of some sort. If it gets shot down, great, it just fulfilled it's job by locating the position of hostile forces. If it actually gets a look at their exact position before going boom, even better.

If it doesn't find anything and runs out of juice, you can just leave it there. Or alternatively collect it if you have some spare time, and then take it back to base to swap in a new set of batteries (or refuel the fuel cells/micro-turbines/whatever).

With the way the necessary electronics is rapidly plunging in price, and becoming more compact as well, well probably see this happening in the not too far future. Hell, the military already have small man-portable UAVs that are "launched" by individual soldiers tossing them into the air. It's really just a matter of time.

[Sea Skimmer]

Exactly. Sorry to burst any bubbles, but at 20 seconds fly time it's not going to get very far. Okay, so they might make a model that flies... uh... 5 minutes on battery power... Just how far is that 20 cm plastic bird going to fly in that time?

Well even 5 minutes would be enough to fly through the main rooms of a modest building, which is the biggest mission we want for really small UAVs. Ground robots already exists for this job, but steep stairs and simple obstacles can stop them too easily to be a satisfactory solution.

[Memnon]

Well, one boost to power that they can manage is to use primary batteries instead of rechargables. They should be small enough that replacing the battery isn't going to be much of a logistics burden, and primary batteries typically have much greater power capacity compared to rechargeables. Example: Lithium-iron batteries have around 3-4 Wh of power capacity in the AA size at the rather high discharge ratings we're probably looking at. Alkaline batteries have more like .8 Wh for the same size.

Hang on, what?
If you're talking about lithium-ion, it's rechargeable. The Lithium-Iron-Phosphate chemistry (LiFePO4) actually has more recharge cycles than Lithium-Cobalt Oxide (the usual, LiCoO), but less power. Alkaline is generally not recharged.

EDIT: I assume you mean LiFePO4 by 'lithium-iron.'

No, I meant the LiFeS2 chemistry used in lithium disposables, like you can get from Energizer. You might be able to get a AA size battery that's lithium ion, but it wouldn't work in anything that took AAs.

I see, I'd forgotten about that chemistry. XD

Maybe if these small birds become easily mass-reproducible, replacing the batteries would be a non-issue since the UAVs themselves would become disposable?

So this would mesh well with primary instead of secondary batteries. I did a little bit of googling on consumer R/C ornithopters, and they seem to be around three hundred dollars for the well-built, study-looking kind. Keeping in mind that these hummingbird ornithopters are much smaller, that should mean they're cheaper (unless the materials are so much more expensive) or about the same cost, ish. That's pretty damn cheap for a UAV - but still much more expensive than grenades (judging by Google, again). The benefits of using one are pretty strong, though - invasion of privacy < getting hit by shrapnel/blinded.

[Starglider]

I did a little bit of googling on consumer R/C ornithopters, and they seem to be around three hundred dollars for the well-built, study-looking kind. Keeping in mind that these hummingbird ornithopters are much smaller, that should mean they're cheaper (unless the materials are so much more expensive) or about the same cost, ish. That's pretty damn cheap for a UAV - but still much more expensive than grenades (judging by Google, again).

R/C ornithopters are barely mass produced, probably a few thousand of the high-end ones a year, though perhaps a hunded of thousand of the toy ones. The military version of this need a camera and a digital datalink, but there's no fundamental reason why that should cost more than a cheap cellphone. I would expect the unit cost to be well under $100 with mass production of >100,000 units, but then there's the need to pay back R&D costs, conduct field trials, the acquisition overhead, devising and delivering user training, plus the standard military markup.

It's going to be interesting to see how much of this unmanned vehicle intel is routed back to intel and command staff. Currently the micro-UAVs seem to be monitored exclusively by soldiers on the ground, but I would expect people back at C&C to start using them (or at least, monitoring the feeds) to get a better picture of what's happening during an engagement. Then again, perhaps this is already happening and I just haven't heard about it.

[Sea Skimmer]

The problem with feeding everything from small UAVs back to the command staff is data overload. How are a couple dozen guys in a command post going to keep track of dozens or hundreds of micro UAV feeds and the context they come from on top of everything else they are already doing? Most of the information will simply not be relevant or useful by the time it is filtered and processed. They've got the bigger picture to deal with, and bigger UAVs to support them. The goal of the US Army is already to supply an entire squadron of Sky Warrior (Army Predator version) drones to every tank and infantry brigade commander by the time acquisition is finished for example.

[PeZook]

The problem with feeding everything from small UAVs back to the command staff is data overload. How are a couple dozen guys in a command post going to keep track of dozens or hundreds of micro UAV feeds and the context they come from on top of everything else they are already doing? Most of the information will simply not be relevant or useful by the time it is filtered and processed. They've got the bigger picture to deal with, and bigger UAVs to support them. The goal of the US Army is already to supply an entire squadron of Sky Warrior (Army Predator version) drones to every tank and infantry brigade commander by the time acquisition is finished for example.

Also, feeding too much data straight to the top of the command chain leads to huge temptation to micromanage. IIRC there were some situations in Iraq and Afghanistan where generals would try to tell officers on the ground where to position their men during a raid.

There's a reason why all human organizations have a tiered command system: the guys at the top simply can't, and shouldn't, deal with all the minutae of tactical decisions.

[Shroom Man 777]

A winged targettable and maneuverable grenade would be pretty damn ingenious. The enemy would have basically nowhere to hide from a directed exploding hummingbird. It'd be like a combination of a grenade and a Quidditch ball from Harry Potter!

[Broomstick]

That is horrifyingly funny - an exploding quidditch ball.

[Hawkwings]

Really? Cause what immediately came to my mind were the seeking Executioner shotgun rounds that the Adeptus Arbites in 40K use.

Hey, there's an idea, have the delivery mechanism be a 40mm grenade launcher tube. If you can solve the problem of extreme g-forces during launch, then you've just significantly extended the range.

[Starglider]

There are a variety of these biomorphic mini-UAV projects around, such as this bat-based one and another with 'perch and stare' technology.

[open_sketchbook]

I wrote a story once where the two world powers go to war using little billions of tiny seeker missiles, which resulted in basically mutual destruction as both sides filled the skies with the buggers and they managed to hunt down most of the world's population. (admittedly, with limited fuel reserves this isn't really an issue, but it was fun to write) The actual main problems I can see with such seeking grenades is target discriminaiton (how is our drone supposed to tell a non-combatant from a target?) and expense (that'd be a really costly weapon for widespread infantry use, especially considering the limited kill-radius you could get out of such a small amount of explosives, with most of the grenade given over to propulsion.)