fix Skydio X2 thrust gear

[christianrauch]

Description

The PR changes the gear of the Skydio X2 such that opposite actuators have the same rotation direction.

Fixes: #219

Checklist

Please check off each item ([x]) once complete, or mark it as [N/A] if it doesn't apply:

• Added your name to CONTRIBUTORS.md (alphabetically by first name)
• Updated CHANGELOG.md:
  • Global changelog (if your change affects the overall repo)
  • Model-specific changelog (if it affects a specific model only)
• Followed the XML formatting/style guidelines (if editing MJCF)
• Ran pytest test/ locally and ensured all tests pass
• Signed the Contributor License Agreement (CLA)

Refer to the contributing guide if you're unsure about any of the steps.

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[aftersomemath]

Yes this fixes the relative pattern of yaw gear's signs so that yaw can be controlled independently of roll.

Did you check that the resulting gears correspond to the right absolute directions? We want to avoid creating a simulated model where every rotor rotates opposite of the real x2.

[christianrauch]

Did you check that the resulting gears correspond to the right absolute directions? We want to avoid creating a simulated model where every rotor rotates opposite of the real x2.

Well, I don't have a real Skydio X2 and I haven't found good documentation that would show the spin direction or propeller pitch to infer the spin direction.

At least you can control the yaw this way.

[aftersomemath]

Yes. Thanks for fixing this. There are many photos of the x2 online that can be used to find the propeller direction.

[christianrauch]

I could only find a good image of the "Skydio 2+" here: https://www.dpreview.com/reviews/skydio-2-review-tracking-tech-wows-but-image-quality-disappoints

For reference, this is the original image:

and this is the model in MuJoCo:

Note that the x-axis is pointing backwards away from the camera and not forward in the direction of the camera.

So the spin direction should be:

Hence, the torque direction for each thruster (CCW > negative torque, CW > positive torque):

• thrust1 -> [-]
• thrust2 -> [+]
• thrust3 -> [-]
• thrust4 -> [+]

as proposed by this PR.

[kevinzakka]

Thank you @christianrauch and @aftersomemath!

[aftersomemath]

Thank you! The model is of the X2, the 2+ was discontinued. However, the propeller directions on the X2 and 2+ are the same, so everything works out the same.

A picture of the X2 can be found here and a copy is included below for posterity.

[kevinzakka]

Thank you @aftersomemath!

[christianrauch]

A picture of the X2 can be found here and a copy is included below for posterity.

I found similar images with folded rotors. But I found it hard/impossible to see the rotor pitch from these images.