Okay thank you

@Zachary-Lowell thanks a lot, will surely try

Thank you for your replies! I will check these links out.

Actually, using docker import fixed it, nvm!

I tried another way to implement this. Currently my flight core is disconnected to the VOXL. I use my laptop as the companion computer, and then I add a ESP Wifi module to Flight Core TELEM1, so my laptop will get the motion capture pose by using vrpn_client_ros, the mavros will transfer the motion capture pose to mavlink data and then send to flight core.

In this case, my EKF2 will converge. The new issue is that the modalAI cannot hover at it's height on flight mode or hold mode, and the modalAI will drift. Also, the mavros terminal will show:"RTT too high for timesync: 78 ms". I am not sure if this is the reason that the modalAI will drifting in position mode.

Any suggestion?

The VOXL 2 Flight Deck exists just for this purpose. You can use that to build your own drone around VOXL 2 to meet your specifications

Hi,

PX4 does not currently support the smooth handling of transitions in/out of stable vision odometry data. We're been working on this with the PX4 team but generally right now PX4's ekf doesn't have any support for handling quality metrics/resets. For both this reason and the fact that the vision data accuracy decreases with distance, we don't really recommend running VIO outdoors. We have done flights outdoors with it, but it requires the right conditions and flying relatively low (and this still will not play nicely with GPS).

@Matt-Turi Yep, turns out switching to wall power fixed it. My battery power was too low. Thanks!

@DarkKnightCH to followup on this -- a reflash of the VOXL Flight Core solved the problem for me. It was in an unresponsive state and I was unable to change parameters or talk to the Flight Core until I reflashed the current release of firmware from ModalAI.

Hi @Alex-Gardner,
We resolved the issue using this forumpost:
https://forum.modalai.com/topic/1050/position-flight-is-not-available-after-firmware-update?_=1658251944919

Thanks for the help

@Isabella-Yu,

If there is no tflite_data pipe, you may be using a version of voxl-tflite-server that does not have the ai_detection_t added yet. This feature was added in v0.2.0, so make sure you are using the latest available version!

@hs775 The instructions here: https://docs.modalai.com/voxl2-px4-build-guide/

are meant to be run on your PC, not the drone.

That depends I guess.

If I merely use the drone for indoor navigation purely by tele-oportation (as opposed to waypoint navigation) then I'm guessing resetting qvio mid flight while moving (and causing a shift in the origin of the reference frame) wouldn't really matter.

Unless I'm mistaken, PX4 uses a velocity controller while moving in Position flight mode. Only when it is stationary does PX4 uses it's position controller (and velocity controller) and that's when a reset of the reference frame origin will really screw things up.

Would love to hear from the devs though

Thanks for the answer, yes I already looked into the code. I was just wondering why the servers timestamp differs from the system timestamp. But then I will change the node accordingly.

@Chad-Sweet No, I didn't try changing any code
Should I try reinstalling voxl-mpa-to-ros ?

Which version of PX4 are you running on the Pixhawk and do you have a magnetometer connected? EKF2 is generally unhappy about about starting without a magnetometer in PX4 1.11 so we patched the PX4 firmware for Flight Core to get around this. In PX4 1.12 and newer the following PX4 parameter can be set to allow EKF2 to allow position mode without a GPS/Mag

## disable mag SYS_HAS_MAG 0 EKF2_MAG_TYPE 5

@modaltb Hi there! May I ask how the loopback test is conducted? How do you get the github file to run on VOXL? Thank you!

@Matthew-Booker will do that, thank you!!