Should Quadrotors All Look Like This?

http://spectrum.ieee.org/automaton/robotics/aerial-robots/iros-2013-should-quadrotors-all-look-like-this#.UoRAb7LvVFs.hackernews

by eguizzo • 13 years ago

228 113 13 years ago

spectrum.ieee.org

speeq• 13 years ago

"The researchers explained that the standard "control" quadrotor had an optimized design, while the triquad was not optimized at all (because its design was constrained to keep it as similar as possible to the control quadrotor). This gave the control aircraft an advantage in power of about 9 percent."

I wonder what the optimized "triquad" would have looked like?

brianbreslin• 13 years ago

Is there a video of this thing in action? I didn't see a link in the article (skimmed it quickly though)

BrandonMarc2• 13 years ago

I can't find one, but these links may help:

Dr Paul Pounds, the project leader:

http://www.itee.uq.edu.au/bme/pounds

http://www.eng.yale.edu/pep5/

Scott Driessens, his student:

http://www.youtube.com/user/sdriessens

krasin• 13 years ago

What software should I use to verify their claims? aerodynamics simulators or how they are called?

shitlord• 13 years ago

Did anyone else watch the linked video of the RC helicopter? It was amazing! I had no idea that those things were so powerful. The only ones I've ever seen are the ones at the mall.

BrandonMarc2• 13 years ago

Agreed. I was impressed by the RC pilot's skill, and noted the craft's behavior was, at times, very insect-like. Or hummingbird.

... and I couldn't help but think, if a life-size helicopter were even capable of such maneuvers, its human occupants would likely die from the G forces.

Clearly such maneuvers seem only possible in an unmanned craft.

h4rrison4• 13 years ago

A cursory glance around the internet suggests that these things are capable of 20G manoeuvres (!), far in excess of the human maximum of 5-10G (depending on axis of acceleration).

http://en.wikipedia.org/wiki/G-force#Human_tolerance_of_g-fo...

eksith• 13 years ago

The mall is more "consumer friendly" ;) The video is from someone who lives RC.

I've only seen one live demo of a "pro" RC pilot doing his thing and saying it was awe-inspiring is a massive understatement. These are folks that eat, drink and breath RC and the amount of time and enthusiasm spent is really astonishing. If anyone can find a way around the laws of physics, it's those in the RC community.

look_lookatme3• 13 years ago

Link

http://www.youtube.com/watch?v=p8t41avFuCc

The thought of a weaponized version of that is terrifying. A weaponized and non-human controlled version is straight up Cyberdyne level stuff...

btown• 13 years ago

Is it possible to scale one of those up to Predator-size? Would the limiting factor be that the power requirements wouldn't scale linearly? Are there properties of the atmosphere that model-scale aircraft can take advantage of that larger-scale ones cannot?

baddox3• 13 years ago

It would be much easier and more effective to just put guns on it.

carterschonwald2• 13 years ago

an 18 year old in the NYC area actually was decapitated by one of those hobbytist copters about 2 months ago. Yes, [one of the reports says] Decapitated.

http://www.nydailynews.com/new-york/queens/teen-killed-remot...

http://nypost.com/2013/09/05/man-decapitated-by-remote-contr...

They're already incredibly dangerous. That trick helicopter just needs one uncontrolled bounce and hit a person and someone will have very very unpleasant day.

Quadricopters seem much easier to make safe afaict.

jotm• 13 years ago

The pilot will be obsolete pretty soon: http://www.youtube.com/watch?v=w2itwFJCgFQ

CraigJPerry3• 13 years ago

I dunno, the payload capacity of these things is shocking until you get to Trex 600 size or above, even at that point it's only "ok", definitely not "good".

At that point (& cost) there are cheaper, more effective (what a cold term) and just down right more terrifying ways, to induce terror.

The small UAVs are destined mainly for aerial photography; they are awesome at that, no question.

rangibaby• 13 years ago

Watching these RC helicopter "tricks" reminded me of the Black Shark helicopter: http://youtu.be/NQier6Fina8

jlgreco2• 13 years ago

First production helicopter with an ejection seat (the rotors have explosive bolts), which is pretty interesting I think.

treeface• 13 years ago

It looked fake to me.

zobzu5• 13 years ago

one of the best/most well known pilot: http://www.youtube.com/watch?v=-us0a6knAKA

spose if you see enough of them it wont look fake no more.

baddox3• 13 years ago

It looks fake when you see it in real life too.

zobzu5• 13 years ago

these things can chop your head off (as per a-month-ago-news-story). and yeah, they're very powerful.

kabouseng3• 13 years ago

Is the helicopter in the video under manual control, or some kind of control system preprogrammed manuever?

jamesaguilar4• 13 years ago

It's under manual control. The tricks are performed by adjusting the trim on the blades to invert lift.

smoyer• 13 years ago

I hope they never give him a license to fly a real helicopter!

zobzu5• 13 years ago

That's kinda stupid garage-scienticism to be honest.

- TFA does not mention the stability issue. Guess why people and drones fly tri, quads, etc and not regular helis. Yeah. (a simplification: stand on one leg. Now stand on 2, and use your 2 hands as well. 4 legs. Stable heh?)

- Yes single large prop is more efficient than many tiny props. But put many large props and the difference isn't so big. Let's take a $1000 ultra efficient quad: www.youtube.com/watch?v=t0SR5bzuFq4 yeah thats 100min of flight time for something smaller than the Align Trex of the video. And it's cheaper too.. The trex flies 5 to 10minutes (6S 5000mah). At equivalent batteries, the trex flies 25min (hint: the trex is not made for efficiency, it's made for 3D). An average camera-less quad flies 10 to 15min (3S 3000mah).

- TFA compares a $5000 Align Trex 700 with a bunch of expensive upgrades - the top of the top - with one very good pilot with decades of training... to a $200 toy.

- $200 toys made of wood and plastic flown by random people are actually pretty good at acro. Why? Because being more stable they're much easier to fly. Ex: http://www.youtube.com/watch?v=qzu5eSZqKpY http://www.youtube.com/watch?v=8QP0QjIsTTM

So yeah. Not exactly impressed when that's coming from a university. I expected much better.

h4rrison4• 13 years ago

Agreed, though the one leg analogy would more accurately be hanging from one arm, since the lift is provided from above.

jlgreco2• 13 years ago

Ehhh... that sounds like an example of the rocket pendulum fallacy: http://en.wikipedia.org/wiki/Pendulum_rocket_fallacy

Maybe there is something more complicated going on with helicopters that I am not aware of, but a helicopter would not be an inherently stable system due to it "hanging" from the rotor because the direction the rotors are providing thrust changes (all other things being equal), when the helicopter starts to tip one direction.

baddox3• 13 years ago

Not to mention simplicity. I suppose it wouldn't matter as much for commercially manufactured rigs, but for hobbyists this thing is a lot more complicated.

joosters• 13 years ago

- Taking cheap shots about their measures of efficiency is hypocritical. TFA says that the researchers did investigate the efficiency and calculated a 25% gain. Presumably their calculations will be in their research paper. You on the other hand have just given a YouTube video and battery capacities as your 'proof'. Heck, the researchers went as far as constructing a device to test their efficiency claims.

- The stability might be an issue for completely manual flight (but then, people fly ordinary helicopters just fine...) but with computer-assisted balancing, the problem can be much reduced. Even existing quad copters often have computer-aided stability.

zobzu5• 13 years ago

I investigated and I say this is wrong. No backup either. = The point.

Now then again, I actually fly every of these so I pretty much know what the efficiency is.

YES they gain 25% efficiency with the tricopter that has a huge prop in the middle, I TRUST that. But the test is meaningless:

If you change the 3 props instead and put big props, with corresponding motors. Guess what. Efficiency is probably higher than the 25% difference. I actually design my own. In fact, I also design my own control boards (the software that is)

Quadcopters use accelerometers, barometers, gyroscopes and magnetometers (and GPS) to ensure stability. Guess what. Measurement from these devices isn't 0ms. Actuating the motor control isn't 0ms.

On regular 1 rotor head heli it isn't either. In fact, it's longer.

Result? it's more stable with 4 props than one prop. It has more inherent stability., even if it wasn't "computer assisted". Timecop did a fully gyro stabilized quad copter recently that shows exactly that, if any "actual proof" was needed. Gyro stabilized can be made with no computer whatsoever (like helicopters are/used to be - that's why they're very, very hard to pilot, real ones and RC ones).

kashkhan2• 13 years ago

The physics makes sense. larger rotors are more efficient because they have lower disk loading. At really low loadings you can lift a human with their own power.

The sikorski prize was claimed a few months ago.

http://www.youtube.com/watch?v=syJq10EQkog

you can get a 600 or 700 class that can do "3D" flying as in the video for much less than 5k. The entry point is below $1k.

their configuration makes sense to get the best features of both. it's not solving all problems but it's an improvement over quads for some tasks.

zobzu5• 13 years ago

the one they're showing in the TFA video is not 1K. The motor, higher servo, blades, etc don't come for free.

kybernetyk2• 13 years ago

I always wondered why there are no "real life" sized quadcopters in use for manned flight. From what I understand the design is simpler and requires less maintenance compared to "normal" helicopters.

Are there any reasons why we're not flying quadcopters around instead?

kabouseng3• 13 years ago

I wouldn't say less maintenance, you've got 4 times as much motors running than a conventional helicopter.

jamesaguilar4• 13 years ago

For the hobby sized quads, it's simple enough to swap a motor. It is definitely easier to maintain since a spare is only 20 bucks and attaches in minutes with zip ties.

kybernetyk2• 13 years ago

I'm by no way an expert on this topic. I just skimmed over the wikipedia article[1] where I read the following:

First, quadcopters do not require mechanical linkages to vary the rotor blade pitch angle as they spin. This simplifies the design and maintenance of the vehicle.

As already mentioned I'm blatantly ignorant about this topic so it would be nice if someone who knows this stuff could chime in :)

[1] http://en.wikipedia.org/wiki/Quadcopter

2bluesc• 13 years ago

Lady Gaga flies with a hexacopter @ http://www.dailymail.co.uk/sciencetech/article-2499954/How-L...

... what the world is coming to?

pygy_• 13 years ago

Morning coffee, this put a big smile on my face.

She (or, probably, her team) has been hacking the entertainment scene for a while.

Dubbing "a dress" what would otherwise be considered a nerd toy is genius. What's not a dress for her?

sanoli3• 13 years ago

She will also fly with Virgin Galactic: http://www.vanityfair.com/online/oscars/2013/11/lady-gaga-fi...

exDM692• 13 years ago

I always wondered why there are no "real life" sized quadcopters in use for manned flight.

The explanation is simple: quadrotor physics do not scale up. The rotational inertia of the rotors is proportional to the length of the rotor blade to the fifth power. Double the rotor blade length and the inertia of that rotor goes up by a factor of 32.

High inertia rotors are not suitable for quadrocopter because fast changes in the rotation of the rotors required for maneuvering them.

Having one big rotor for lift and three (or more) small ones for control is a very good idea. Another idea that is being researched is having multiple quadrotor copters flying in co-operation.

rpenm• 13 years ago

The moment of inertia for a uniform rod pivoting on end is (1/3)ML^2 where M is mass and L is length.

Assuming mass scales with length, inertia should be proportional to the third power. Where am I wrong?

kashkhan2• 13 years ago

there are :) human powered at least:

http://www.youtube.com/watch?v=syJq10EQkog

Zigurd2• 13 years ago

There was some crazy man/scammer building a manned flying thing with six snowmobile engines powering six rotors. I remember this because I was working for Ray Ozzie at the time and he was telling people in the office about it because he had invested in it.

rsl7• 13 years ago

Moller?

bambax2• 13 years ago

It takes (at least) six months to learn to fly a helicopter (or, in the case of the linked video from the article, probably a lifetime).

It takes ten minutes to learn to fly a quadcopter.

nakedrobot2• 13 years ago

No. Try a cheap quadcopter without any auto stabilization. it takes more than 10 minutes. Maybe 8-12 hours of flying time to become proficient enough to, say, risk mounting something breakable (like a camera) onto it.

Less time than a normal helicopter though, so I get your point.

bambax2• 13 years ago

I have two quadcopters -- that both have auto stabilization: Parrot AR Drone and DJI Phantom, and it did take around 10 minutes to learn to fly them.

Is it even possible to fly quadcopters without stabilization? Would they not start spinning out of control as soon as they take off?

marvin• 13 years ago

This is really cool, and probably a better solution than quadcopters for many applications. However, what we would ideally want in the end, would be a tilt-rotor of some sort which incorporates wings. Fixed wings are a lot more efficient than active lift, and is necessary if you want to squeeze more speed and range out of your UAV.

Pxtl2• 13 years ago

So, like, a remote-controlled Osprey?

neya2• 13 years ago

Wow, readers, please do yourselves a favor and watch the linked video in that article! [https://www.youtube.com/watch?v=p8t41avFuCc] though it's slightly WTf-y, it's definitely cool. I'm just wondering - Are heli's better at carrying my Canon DSLR or would the triquad work better for this?? Anyone with experience, please suggest, as I am SO gonna build this tri quad (and integrate it with my Raspberry Pi) :D

fijter• 13 years ago

A heli carrying a DSLR is big, expensive, hard to fly, dangerous and maintenance intensive. It's good for lifting a DSLR but I would go with a multicopter, the more blades the better. A hexacopter or even an octocopter are very stable, easier to fly, possibly cheaper to build and have a fail-safe, if a motor fails you can land it safely because the other motors take over. You can find a nice video about the subject here:

http://flitetest.com/articles/T_Rex_800E_DSLR_Camera_Gimbal

It demo's a Heli carying a DSLR and afterwards (after the first 7 minutes of video) they talk about this compared to multi-rotors. For the best video results you should probably pick the stable octocopter, you can fly FPV as well with a multicopter, great for live previewing your video.

neya2• 13 years ago

Thanks! That's pretty informative! :)

Dave_Rosenthal2• 13 years ago

For any interested HN'ers:

Though quads are much more trendy in the tech world today, as you can see from that video, plain-old RC helicopters are pretty amazing today. The biggest advance in the hobby IMHO has been the development of very small (~1oz) RC helicopters (e.g. [http://www.bladehelis.com/Products/Default.aspx?ProdID=BLH33...) that include the same fundamental control system (called collective-pitch) that enables these large (~10 lbs) helicopters to be so acrobatic. The size shrink brings huge benefits: Small helis basically don't get damaged in most crashes, they mostly can't hurt you, and they are cheap.

Today's super-stabilized and gps-locked quads are like a modern AAA game: you press "A" over and over and watch them go, and say "Wow!". A micro collective-pitch heli is like an old-school game--you're going to need a lot of restarts, but is great fun and will provide many hours/months of challenge.

CraigJPerry3• 13 years ago

How does this perform cross wind compared to a traditional micro quad?

That main rotor will act like an umbrella does in a gust of wind. The micro quads don't suffer the same problem.

I'm not sure of the physics involved (and if anyone knows i'd love to learn more!) but just from hands-on piloting both micro quads and larger helis, the micro quads are some of the least susceptible to cross winds i've ever flown.

loup-vaillant2• 13 years ago

Accelerometers are bound to detect such adverse effects. Correcting roll and pitch automatically shouldn't be hard. Also, I would guess that the higher the main rotor (compared to the centre of mass), the more susceptible the device. So, if you build the main rotor as close to possible as the centre of mass, you should have a more neutral design.

While we're at it, I've flown a little toy with two main rotors, rotating in opposite directions, controlling power and yaw, while the tail rotor controlled pitch (the whole thing was auto-stable, so there was no need for roll control). It worked like a charm. So, if we're willing to suffer the complexity of an actual swashplate[1], we could try a design with 2 main rotors, one above, and one below, hereby getting a neutral, manoeuvrable, and (hopefully) efficient design.

[1] https://en.wikipedia.org/wiki/Swashplate_%28helicopter%29

CraigJPerry3• 13 years ago

AIUI there's all sorts of considerations & trade-offs. E.g. the motors have a limit to how quickly they can change output (up to around 400Hz today).

Because there's no swashplate on the upper rotor of those wee co-axial rotor designs, they're really ineffective in cross wind. You can't apply the required corrective input when the top rotor gets blown off axis.

Dave_Rosenthal2• 13 years ago

The conventional wisdom is that rc quads are _more_ susceptible to winds than rc acrobatic helicopters (and have lower max speeds) largely because they usually have a "cupped" shape to their props (to maximize efficient static thrust) which can catch catch a cross wind. In contrast an acrobatic rc heli has perfectly straight/flat blades that change pitch to create lift and are quite "invisible" from the side.

That said, in this design, having two different types of rotors mounted non-symmetrically seems bound to induce some yaw from winds.

rplst8• 13 years ago

I for one welcome our new...

jwr2• 13 years ago

I really like this, because it shows that quadrotor (and flying vessel in general) research is far from "done".

Most people would tell you that you'd be insane to try to design new types of aircraft, because we already know how to do that. This research proves otherwise.

Gravityloss• 13 years ago

One might almost say that anybody knowing basics of Newtonian mechanics would say maximizing your blade swept area reduces power needed to stay aloft. Then there's stuff like tip vortices etc. So just one big rotor is great.

The problem becomes control. With many small rotors you don't need to control blade pitch, as the electric motors have high torque and the blades are light, so you can control with quick engine thrust changes. And it's that what makes the quad rotor system much more attractive than a normal helo. The traditional helicopter pitch control system is complex and requires precise high stress components that are well built and maintained.

But it's a very inelegant brute force system in my opinion. The pitch control could be improved a lot as well if you have powerful electromagnetics with very precise control.

You still need a tail rotor though.

The real helicopter is seeing some real progress at the moment as well, all in the name of higher speeds. There's Sikorsky's X2 and AVX's modified Kiowa. They have coaxial main rotors with a pushing tail rotor and two pushing fans respectively. Then there's tilt rotors by Bell and Boeing, IIRC. These were, at least at some point, all part of a US scout helicopter competition. And then there's the Eurocopter X3, with no tail rotor but twin pulling rotors like a twin prop airplane. X2 and X3 have flown.

Should Quadrotors All Look Like This?

Related Stories