DuoCopter | Hacker News Reader

DuoCopter

research-drone.com

180 pointsbased241 comments

Here is a somewhat similar design, from 2016. It uses gimballed rotors at either end instead of modulating the angular velocity within a rotation cycle. Seems more sensible to me to do it that way, and it shows fully controlled flight (even flips it over at one point) instead of the little hop in TFA.

https://www.youtube.com/watch?v=b14_XQ5nd2A

Really curious how that thing generates torque since it doesn't have asymmetric props.

edit: I'm dumb, it has gimbaled props. That makes OP's link much more impressive, since that thing doesn't require gimballing which is expensive.

yea there is another top level comment by gtirloni which links to a different example which does it without a gimbal, by applying a sinusoidal current with variable amplitude and phase to the motors. They have symmetric rotors though, so your question applies there! I wonder if maybe the rotors are in fact slightly asymmetric and just look symmetrical.

edit: I got it, this is super neat: The blades are hinged at a 45 degree angle, but in opposite directions. See Figure 1 and 2 here

http://www.modlabupenn.org/wp-content/uploads/paulos_an_unde...

Yeah, that still requires moving parts on the rotor. OPs design requires no additional moving parts. I like it.

> Seems more sensible to me to do it that way,

The interesting part to me of the OP's design is it has only two identical moving parts.

Adding 2 * 2 DOF gimbals is four more actuators to go wrong, for a total of 6. That flight demo is pretty nice, though.

Opposing rotors is a proven helicopter design, https://en.wikipedia.org/wiki/Coaxial_rotors

By placing the rotor under the vehicle you remove the mechanical complexity of the rotor but you don't have any way to land... Unless your landing gear is as wide as the rotor and extending around it, or some kind of arm catching the whole vehicle from the side.

Interesting. Watching it move, I wonder if copters like this would be better suited for things like navigating through trees and forests. Small quads can do it too but these should be able to do it while carrying larger payloads.

Re: Patents. My understanding is that it doesn't need to be "revolutionary" to be patentable.

> My understanding is that it doesn't need to be "revolutionary" to be patentable.

How would one measure revolutionaryness anyway? If it hasn't been done before and isn't just a variation on an existing thing (e.g. sports car but in green), that sounds like a good enough thing to award a patent for.

Which is not to say that I agree about all the details, e.g. that software should or shouldn't be patentable, or how long a patent should be valid, etc. Just that one can get a patent for something that's not "revolutionary" makes sense.

Thomas Jefferson would disagree with you strongly, he was quite clear that he believed patents should require a very strong degree of novelty and differentiation:

See, for example:

http://www.let.rug.nl/usa/presidents/thomas-jefferson/letter...

Sadly we've strayed very far away from how Jefferson thought patents should work in this age of "This thing... but on a computer" patents.

I like this design better. The mechanical linkage isn't complicated. Balanced two-bladed rotors. You're only sending minor impulses at certain points in the rotation rather than changing the speed significantly.

They have another design that can go from 0km/hr to 100km/hr in 1.5 seconds: http://www.research-drone.com/en/extreme_climb_rate.html

What is not proven for me is that this is more cost effective than the quadcopter. The unbalanced rotors look like they will need expensive bearings. The complex speed control looks like it is going to be very expensive by the time they have tackled the efficiency issues. This is vs using just two more identical motors. This brings economies of scale, plus you can use smaller motors. The efficiency issue alone may mean a quadcopter needs less battery capacity, allowing a lighter payload. Always interested in a novel design, but still lots to prove

This is a cool design. I think the group from University of Maryland developed a single motor single actuator drone back in 2009, here is a link to video...

https://m.youtube.com/watch?v=u23Hqq8QbeE

And article about students involved...

http://eng.umd.edu/news/story/maple-seeds-inspire-robotic-fl...

It appears to rely on slowing down and speeding up the rotation within a single cycle. I imagine that would be quite wasteful, unless there is some kind of regenerative process involved, ie dumping the power into a capacitor and then releasing it again, or something like that. Does anyone know how one would do that?

FTA:

"By means of a further developed motor controller, the battery can be charged efficiently with the brake energy. This works with high efficiency as with a flywheel. For acceleration, energy is removed from the battery."

Interesting. I would have thought the battery is not the right place to store that energy, as it will be literally doing thousands of cycles per minute...

Yeah, sounds like a job for a capacitor. Still, you're going to have lots of losses vs just not using the extra energy in the first place. Why not do something like a swashplate ( https://en.m.wikipedia.org/wiki/Swashplate_(aeronautics) ) like helicopters? Instead of varying the speed of the blades, just vary the pitch of the blade during rotation. But then you've just invented the helicopter. Use counter-rotating blades if you really want to eliminate the tail boom, see https://aviation.stackexchange.com/questions/8642/how-can-a-... for some example prior art).

My first thought when seeing this is also around the longevity of those poor bearings, given the eccentric load. I wonder how much this would benefit from a motor with many more poles (to control the rotational velocity to a finer degree) and the lightest props you could find (CF is already used extensively in RC help blades). I also wonder if it'd be possible to use a horizontally hinged blade which could shoot forward + rock backward in a controllable fashion to control how much lift you get during a particular phase of rotation. I'm sure this is not a new idea.

The whole point is to avoid the complexity of swash plates and similar mechanisms. This is just two rigidly mounted motors with asymmetric props.

Isn't this a job for a mechanical system, like a rotor mounted off axis with a counterweight?

Now you have to move the counterweight in and out, which gives you the same problem.

Why are there so many claims without even a demonstration of them?

Because the guy is trying to sell a license to the idea to fund development of those other claims.

And unless this guys is Evan Ulrich[1] who explored the space[2] of single bladed propelleor microscale UAVs back in the 2010-2012 time frame I expect his patents will have to require the UMD patents as well (making licensing more complex)

[1] http://www.avl.umd.edu/people/evan-ulrich.html

[2] http://www.dcsc.tudelft.nl/~bdeschutter/private_20100705_acc...

What do you mean exactly ? There is a video embedded with the drone in action. What claims do you imply ?

The site and video claim all kinds of performance benefits, a better video picture, longer flight range, higher efficiency, quieter, etc. The video shows it flying about 5 feet and then a bunch of motion graphics. There's no substance to this design. If all the claims are true, I'd love to see it in action and learn more. If you go by just the video, it can barely get off the ground. Not so innovative when you look at it like that.

Well, the video shows a drone flying and some concept studies nothing more. No close up of the drone, no targeted flight in a certain direction, nothing. It lifts off, moves (seemingly uncontrollable) and lands again. What claims can be made from this about the used technology?

Nice. What about vibrations?

I understand that the author is trying to sell us the product and get it developed but what are the cons of using a duocopter vs your normal quad? I generally think of decisions as tradeoffs and want to know the cons as well when considering an option.

Note that I'm just an drone enthusiast in that I buy and use them but I do not develop or go DIY which is why I ask this question.

As a fellow enthusiast, this doesn't look particularly agile (or fun).

> but what are the cons of using a duocopter vs your normal quad?

Higher lifecycle cost due to lower fatigue life resulting from higher cyclic loads on vital components such as rotors.

Cheaper, more dispensable drones that can be deployed in larger numbers in dangerous environments?

It's a flashy ad designed to trick stupid, greedy VCs.

Cool.

I struggle to think of a scenario where this would be a better fit that a quad-copter.

Maybe where the with of the aircraft is a major problem, in some sort of tunnel or suer.

>I struggle to think of a scenario where this would be a better fit that a quad-copter.

single-engine controllable aircraft is one example in the demo video. They are showing off the algorithm; not the platform.

Patent protected and not revolutionary.

>not revolutionary

Well it does revolve so...

    >Inexpensive

Consumer-wise or manufactor-wise? I'm thinking the latter

that 1 second of "stable hovering" really sold me.

(spoiler: it didn't really look that stable)

Is it just me or does this seem to be a scam? My BS detector is going off every second of this video.

I have a feeling the inventor is looking for cash with no substantiated product built.

> Is it just me or does this seem to be a scam?

My thoughts exactly. It's such a long video but it essentially consists of a series of crude animations peppered with very short and unsubstantial footage of what appears to be an implementation shown afar doing practically nothing, and even so while displaying a warning message as it's just a prototype.

This does look like a half-baked scam.

the website and even the domain name are suspect.

Reply on news.ycombinator.com