Rhys Mainwaring

Looks like it's hardcoded to use the travelling wave version: https://github.com/srmainwaring/asv_wave_sim/blob/40eed5158fcab37421826ffa3ba33aab94d55e28/gz-waves/src/LinearRandomFFTWaveSimulationImpl.hh#L213 So, `Cos2s` always.

Hi @ConnorDTaylor, the wind speed is a parameter of the wave spectrum function. The resulting spectrum is sampled for the amplitudes which are then feed through the 2D inverse FFT...

I'm afraid the modelling of how a sea state builds is an entirely different (and very much more complicated) problem. From my understanding the steady state spectrum and spreading functions...

> There's a stalled PR #2336 that was fixing the units. There we decided to add a flag to enable this so as to not break behavior. Can we do...

> Shouldn't NED be the default then? The reasoning is that for Gazebo the world frame is ENU, so to be consistent with the orientation given by other sensors (say...

Sorry, centre at `Vector(0, 0, 0)`. These are the normalised position vectors rather than the coordinate vectors. I noticed this when using the `PathSegments` to generate waypoints. I was calculating...

Sure, let me take a look and I'll see if I can provide an example that illustrates the case. Although the easiest way to replicate the issue is with the...

@Jaeyoung-Lim here's the example: https://github.com/srmainwaring/terrain-navigation/tree/prs/prs-path-segment-tests Suggested fix in: https://github.com/ethz-asl/terrain-navigation/commit/b7b3f378e0b2ae6c390df81624c2e1617a1a1502 (Needs cleanup and debug prints removing) If you agree with the suggested approach I'll tidy up and submit a PR, should...

Thanks for the PR @snktshrma. Looks good. Have suggested a few changes. A screenshots of the QGC settings page might help clarify the setup, but not essential.

Adding resized image here so it can be referenced from as an asset in the main repo.