James Jackson

Yeah, we should be able to handle a disturbance like that. I really like the notch filtering idea. This might be a good dataset to use to develop it.

Awesome. Thanks so much. I'll hopefully get some time to look at this tonight. In your testing, did you notice if the bias error always went the same way?

Ultimately, I believe reason that the filter isn't working very well is that not very many of the IMU measurements are making it into the filter. (check out lines [156-157...

Haha! oops. Yeah, my bad. I filtered the IMU messages in my rosbag parser and things look a lot better now. I did a little bit of tuning ``` ACC_LPF_ALPHA:...

I think the cleanest way to do this is to publish a `status` message from fcu_io. This could include things like whether or not the safety pilot is in control,...

I just added a status message which is published by `fcu_io`. This includes a flag of whether or not the flight controller is in failsafe mode. You'll have to watch...

The latest `status_message` branch should allow for onboard commands to be accepted even if RC is lost.

When you get around to testing this, I can show you a quick hack that tricks ROSflight into thinking it's lost RC when you flip a switch.

We should just make this a parameter so it can be turned on, but by default is turned off.

@dpkoch, could this be related to the issue that we've seen that every time we change any parameter, `rosflight_io` responds that the `ARM_CHANNEL` has a new value (even if it...