Robot ostrich spy outpaces world’s fastest sprinters

Researchers are building a prototype design that would enable a two-legged robot modeled on an ostrich to run at around 50mph. In simulations, it can already outpace human sprinters. Dubbed FastRunner – presumably to counter any “neep, neep” Road Runner jokes – the robot is being developed by teams at the Florida Institute for …

[bobdobbs]

umm....

shouldn't the title read:

"boffins hope, pray, and speculate that their dreams of robot ostriches will sometime in the far future be realised"

I don't see any evidence that this this is anywhere close to existing, or in fact that they have anything "done" besides a computer model and animation of two funny looking legs "running" while suspended in mid air.

[PaulW]

Hey - they have 70% of one leg done.. which must count for something? Other than perhaps a robotic supersized drumstick for Chrimbo?

[Anonymous Coward]

But. But...

You didn't specify you wanted it to work in THREE dimensions...

[James Micallef]

In simulations, it can already outpace human sprinters.

Was the simulated human modeled after it's designers?

[Anonymous Coward]

Tripwire...

Boom!

[It wasnt me]

Indeed

Exactly the point in fact. Much better a robo-turkey gets its goose cooked by a trip triggered bomb than a human being / american soldier.

[Boothy]

With cameras (multiple wave lengths), and ultra sonic scanning, it would probably notice a trip wire a lot easier that a human would.

I can see these being used as the new 'Point-man', with the squad watching via a tablet or head up display some ways off, and a mic/speaker in case they bump into a local.

[laird cummings]

Robot overlords...

Can't beat 'em, and now, can't outrun 'em, either.

[Martin Huizing]

Beep beep...

Now the North Koreans only need to build a Wile E. Coyote robot to chase this one.

Beep beep!

[Graham Marsden]

So is someone going to develop...

... a robot coyote to chase it?

Hmm, nah, it would be doomed to failure from the start, especially if it was built by Acme Corporation!

[Anonymous Coward]

Haven't you seen the (allegedly spoofed) episode where Roadrunner was killed?

[Anonymous Coward]

as robot wars has taught us...

a robot without a SRiMech is a non starter.

[BorkedAgain]

Someone's been playing Metal Gear...

...while mashed on 'shrooms. Fine if you're a student; worst case scenario your end-of-term paper is late and poorly researched. It won't be the only one. But if you're in charge of military research somewhere...

[BorkedAgain]

Downvote? Really?

I'm just trying to work out who I could have offended with that one. I'm guessing either a student with an overdue paper, an arms merchant who's worried his dirty little secret is out or a pixie who wants to discourage the wholesale consumption of his housing stock.

Which was it? Go on, come clean...

[Anonymous Coward]

Oh great!!

So I guess these robot designers have never seen either the Terminator OR Jurassic Park movies? Otherwise they wouldn't be building robotic velociraptors programmed to stalk puny humans!!

[James 63]

Je suis sceptical...

Evolution has had quite a few goes at getting ostrich legs just right. Less efficient designs get eaten. I'm surprised these peeps think they can just study some video and "devise a more efficient mechanical system". Ostriches will have muscles in their lower leg for a very good reason, which so far seems to have eluded this crowd.

And as for "In simulations, it can already outpace human sprinters", well, in simulations I can land a 747 on an aircraft carrier. Come back when you've actually got something to tell us...

[Anonymous Coward]

No, you cannot

I get the point you were trying to make in that simulations cannot accurately model the real world perfectly but you do your argument harm by using a bloody silly 'I can land a 747 on a carrier in a simulation' analogy.

Everyone with half a working brain knows that that would be impossible in any reasonably realistic simulation and we are giving the benefit of the doubt to the scientists that they are using a relatively accurate simulation software rather than 'Looney Toons Physics simulator V1.4'

[Anonymous Coward]

dude, where can I get my hands on V1.4 of the Looney Toons Physics Simulator?! That would be awesome!

[Fuzz]

I suggest you upgrade to V1.5 it fixes a bad anvil related issue.

[TeeCee]

Looney Toons Physics?

Not necessarily. A US fleet carrier can steam into the wind at about 30 knots. Now, the stalling speed of an unladen 747-400 with full flaps and pretty much bingo fuel is 115 knots. That's a net approach speed of 85 knots. Now, if we assume a good, solid 70mph gale over the deck, that's 60 of those knots dealt with, leaving a mere 25 to be dealt with by reverse thrust and braking in the length of the deck. Perfectly feasible.

Of course the wing would probably hit the island and there's absolutely bugger all margin for error, so you wouldn't want to do it, but it's quite possible.

[It wasnt me]

Idiot

As you point out in your own post its only 'quite possible' if the simulator ignores the fact that the plane has wings.

To re-assert the previous poster, in any reasonable simulator it is just not possible.

[Anonymous Coward]

They landed a C-130

The US Navy DID actually land a C-130 on a carrier. You can see videos and pictures after a quick google.

The C-130 has a wingspan of about 132ft and JUST managed to fit without decapitating the carrier and suddenly leaving a lot of promotion prospects for the lower ranks.

A 747-400 has a wingspan of 211 ft....

I also doubt, but cannot prove, that the 747 wing height off the ground is not enough to go over the tower.

[Anonymous Coward]

the selective pressures are different.

it doesnt have to be more agile than a predator, or be constrained by muscle performance, or be limited in energy use by how much food an animal can find/eat.

i assume:

it has to travel in human sized environments, over rough terrain, at a speed comparable to humans.

perfect is the enemy of good.

[Richard Scratcher]

@James 63 - Devising a more efficient mechanical system

It took hundreds of millions of years for evolution and natural selection to come up with the ostrich leg but it's not that difficult to come up with a more efficient mechanical system.

Evolution produced an ostrich that has to be able to grow from a tiny fertilised ovum into an adult bird. It also has to be able to repair itself for damage and wear. The power source has to come from digested food. The mechanical leg is free of those requirements,

Gentlemen, we have the technology. We have the capability to build the world's first robot ostrich. Better than it was before. Better, stronger, faster.

[fajensen]

.... With the better, stronger, faster "GashMaster(tm)" Homeotropic Chainsaw Attachment!

[Richard Scratcher]

Nah, four legs good, two legs bad.

Boston Dynamics "Big Dog"

http://www.youtube.com/watch?v=cHJJQ0zNNOM

No! Down boy! Bad Big Dog!

[Anonymous Coward]

Robotic Ostrich eh?

Does it bury its head in the sand at the sound of gunfire?

[Neil Barnes]

I seem to recall ostriches have one hell of a kick

Can you imagine how *embarrassing* it would be at the enemy field hospital?

"You were kicked by a robot bird? No medals for you, Sunny Jim!"

[JDX]

re:Je suis sceptical

Human amputees can run pretty well on replacement legs that are essentially a single piece of spring-steel.

A real bird has to be long-lasting. A robot doesn't need to run for 10 years without wearing out.

Also, some systems that are efficient, can be ill-suited to evolutional processes if there isn't a way to get there through many gradual iterations. So planning the final result, it's possible to create something evolution would never have got to.

[Ralph B]

Unexpected Expectations

"computer simulations show no reason why is should not exceed expectations."

Regrettably, they cannot expect it to exceed expectations at the moment, they can only expect it to meet expectations. That's what expectations are. Later, of course, when it is all over, they may look back and realise that it did indeed exceed expectations. That'll be nice for them then. But not now. It's just too soon for that now.

Unless they have a time machine.

[Asgard]

Here's some more details on how it moves. (Its a very interesting design but I think the 50MPH is more than a bit speculative at this stage). At this stage of design in the following video, it reminds me very much of a Theo Jansen style of leg design...

http://www.youtube.com/watch?v=-_k9SZAX8GY

(If you've never heard of Theo Jansen, look for his amazing work on Youtube. If you search for "Theo Jansen mechanism", you will find animated diagrams of how it works).

[t_lark]

Simulations are used to do the normal engineering math that is impossible to do by empirical trial and error. Complaining simulation has been used to design the parts is like saying we shouldn't use equations to build bridges.

Look at some state of the art style passive walkers http://techland.time.com/2011/10/26/watch-a-robot-that-can-walk-without-motors-or-electricity/

Computers are used to make sure the legs are balanced and will operate passively (or in this case use one motor to drive many joints), which was highlighted as an innovation in this work.

[Craig 2]

Any two-legged robot is doomed to failure unless it can get itself up after falling over.

[Cam 2]

Exactly

I wondered why no-one had pointed this out. How difficult would it be to trip or topple a two legged device, how much engineering would have to go into making it self-righting?

[JulianB]

"computer simulations show no reason why is should not exceed expectations"

Does that actually mean anything at all? (even ignoring the typo)

If it does, it's probably something like "Based on the computer simulation, we expect it should do X well, but we don't trust the simulation, so it might be better, or indeed worse"

[Imsimil Berati-Lahn]

It's not an ostrich

It's a tauntaun.

[earl grey]

robot ostrich?

Is it better than robot chicken?

[Anonymous Coward]

Pfft, I have a better ostrich simulation...

http://www.youtube.com/watch?v=xWIQtmwkAO0