I've seen a few photos taken by Ingenuity of its own shadow, and the angle between the two sets of rotors is different in each.

That suggests to me that spinning the counter-propagating rotors at very slightly different speeds could make the aircraft turn via some conservation of angular momentum argument; the two torques wouldn't exactly cancel.

But the two torques could also differ by setting the angles of attack of the two rotors slightly differently as well.

Question: Does Ingenuity rotate via differential rotor speeds or differential angles of attack?


Hackaday has a well-written article describing nicely the control system for Ingenuity.

Another way in which Ingenuity differs from terrestrial multicopters is in the flight control systems. Where most quads only have fixed-pitch rotor blades and use differential motor speed to achieve pitch, yaw, and roll control, Ingenuity uses a pair of swashplates to control each rotor’s collective and cyclic pitch. Each titanium swashplate is controlled by a trio of tiny servos anchored to the rotor mast, and is connected to the rotor using connecting rods machined from polyetheretherketone (PEEK) plastic.

ingenuity control system

Image courtesy of linked site.

I'm quite surprised to see swashplate controls on such a small 'copter, but the biggest advantage, I think would be in the precise control such a system would provide. A bit more complexity, of course, but the parts are not so large as to impact the craft, as the recent flights would attest.

  • 1
    $\begingroup$ Thanks! Can we say definitively then that Ingenuity rotates by differential angles of attack and not via differential rotor speeds? $\endgroup$
    – uhoh
    Apr 23 at 1:02

As @fred_dot_u's answer shows from a blog post, the control mechanism is two swashplates.

The manufacturer of the six motors that control the swashplates confirms that:

Six DCX precision micro motors with a diameter of 10 millimeters are responsible for moving the swashplate and hence adjusting the inclination of the rotor blades - i.e. for controlling the vehicle.


A swashplate tilts and translates (up/down; collective pitch), which changes the pitch of the blades. Opposite collective pitch with two rotors results in yaw, as used in coaxial Earth helicopters (dissymmetry of torque).

But what about the two propulsion motors?

They are both run at the same speed:

The vehicle features a total of eight motors: two brushless direct-drive propulsion motors driving the two rotors, and six brushed servo motors for controlling blade pitch via a swashplate on each rotor. The propulsion motors are controlled to keep the rotor speed constant, at a setpoint chosen depending on the atmospheric density at the time of flight. The motion of the vehicle is controlled by modulating the blade pitch, which affects the amount of lift and drag produced by the blade. The swashplate mechanism allows for two types of blade pitch control: collective control changes the average pitch over a rotation, and cyclic control allows for periodic modulation of the blade pitch at a frequency of once per revolution, with a specified phase and magnitude. [emphasis mine]

— Grip, Håvard F., et al. "Flight control system for NASA's Mars helicopter." AIAA Scitech 2019 Forum. 2019. (PDF; nasa.gov; free access)

Control inputs would result in different loads on each rotor; the lightest way to make both rotors run at the same speed is independent motors, instead of a larger motor running both rotors with a gearing for the counter-rotation.


[Edit: asdfex pointed out in a comment that they do have two motor systems and the two sets of blades therefore are not geared together. So it's possible that they do do tricks with speeding up & slowing down the blades, and my argument that two motor systems are risky and heavy is clearly not correct. I'm leaving the rest of this here as I still believe that they use fairly conventional helicopter control systems (and certainly they use varying pitch), although clearly it's more complex than I thought.]

It's important to realise that Ingenuity is a helicopter: it's tempting to think of it as being a bit like the kind of multirotor drones we all know about, but it's not, at all.

With a helicopter with counter-rotating blades like Ingenuity you could indeed turn by varying the speed of the blades independently of each other and relying on angular momentum conservation. (A helicopter without counter-rotating blades do that because if it speeds up and slows down the rotational speed of the blades it also gains or loses lift. It could probably turn by controlling the rotational speed of the tail rotor I suppose.)

But using this technique to turn is absolutely horrible. It's horrible because it means that you need to be able to chose the rotational speed of the blades independently of each other. You can't do this with gearing from a single motor, so you need two complete motor systems to do this. That's both mass and really nasty failure modes: if anything happens which causes one motor to slow down you have a catastrophic accident rather than descending to the ground in a controlled way. You don't want even the possibility that the two rotors can rotate at different speeds.

And it's also horrible for another reason: as I said at the top, Ingenuity is a helicopter, and helicopters, to be useful, must have pitch control of their blades. Without it all you can do is go up and down, turn (at horrible cost in mass and safety) and ... let the wind blow you where it will. That's a very expensive, very noisy balloon (well, balloons can't turn).

So, in order to be able to move about with a helicopter, you need to vary the pitch of the blades both 'collectively' (increasing or decreasing the pitch by some constant amount) and 'cyclically' (varying the pitch depending on the angle of the blades. The first of these lets you control the overall lift, and the second lets you move around. Ingenuity does both.

  • 1
    $\begingroup$ I just had a breakthrough, profound realization; since the blades are counter-rotating, of course in different photos they will have a different angle with respect to each other. Somewhere in a recent comment I saw a link to this video and I guess the image stuck with me. $\endgroup$
    – uhoh
    Apr 23 at 11:43
  • $\begingroup$ According to detailed drawings, there are actually two separate motors, one for each set of blades, in the cylindrical boxes right above them $\endgroup$
    – asdfex
    Apr 23 at 14:30
  • $\begingroup$ @asdfex: yes, that is true, isn't it! I didn't know & I'm surprised by that, as it seems to me that introduces a nasty vulnerability if one motor fails. Perhaps it was just enough lighter than the gearbox they'd otherwise need (and which it looks as if an earlier prototype had). I will update my answer, as it may be wrong now. $\endgroup$
    – user21103
    Apr 23 at 15:27
  • $\begingroup$ @asdfex do you have a link to those? I'd love to see them. $\endgroup$ Apr 23 at 16:09
  • $\begingroup$ @OrganicMarble Here you are: rotorcraft.arc.nasa.gov/Publications/files/… Figure 7 is the most detailed view $\endgroup$
    – asdfex
    Apr 23 at 16:42

It yaws (rotates about a vertical axis) using differential angles of attack.

I found the answer by looking at the paper linked to in ymb1's answer. Thanks to ymb1 for pointing me in the right direction! (No pun intended.)

That paper says:

Yaw (heading) is controlled approximately by antisymmetric collective, which is defined as one-half the difference between lower and upper collective. (p. 15)

— Grip, Håvard F., et al. "Flight control system for NASA's Mars helicopter." AIAA Scitech 2019 Forum. 2019. (PDF; nasa.gov; free access)

In other words, yaw is achieved by setting one of the rotors to a finer pitch (a shallower angle of attack), causing that rotor to produce less drag, and setting the other rotor to a coarser pitch (a steeper angle of attack), causing that rotor to produce more drag.

A later paragraph on the same page explains how the rotor speeds are kept equal to each other during such a maneuver. Normally, setting one rotor to a coarser pitch (and thereby causing that rotor to produce more drag) would thereby make that rotor slow down, which would cause control problems. However,

This issue is solved by the addition of a feedforward path from the collective control to the motor input voltages, which are used to control the torque applied to the motors. When actuating the collective control, the voltages are modulated in anticipation of the change in rotor drag, thus nullifying the effect of the collective input change on the rotor speeds. (ibid., p. 15)

  • $\begingroup$ Thank you for your in-depth ad well-sourced answer! Btw there's another aviation-like question that may still need one more answer: Could an aircraft ever simulate Martian gravity perpendicular to the aircraft's floor? $\endgroup$
    – uhoh
    Apr 23 at 22:03
  • $\begingroup$ Shouldn't a "sharper angle of attack" be one closer to 0? I think "larger angle of attack" would be more correct, no? $\endgroup$
    – Jens
    Apr 24 at 13:14
  • $\begingroup$ @Jens Good point. I think the word I was really looking for was "steeper." $\endgroup$ Apr 24 at 17:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.