@tom I'm not having the same issue with GPS lock. I can get a GPS lock, but my GPS is showing as disabled in QGC. Is there a PX4 parameter loaded on the Sentinel's that defaults them to disable GPS? Changing the EKF2_AID_MASK to enable GPS doesn't due anything. *Ignore this part, just didn't have COM_ARM_WO_GPS set to 0 Also, I'm still concerned about the constant MAG#0 timeout errors, the unstable/unrealistic position estimates I'm getting for Local Position NED in QGC and the high loss rate I'm getting over mavlink. Do you have any suggestions for troubleshooting and is there a way to synchronize the clock on the voxl2 with the clock on the GSC? Thanks

@Eric-Katzfey this is our GPS status right now INFO [muorb] SLPI: qshell gotten: gps status INFO [muorb] SLPI: Main GPS INFO [muorb] SLPI: protocol: UBX INFO [muorb] SLPI: status: NOT OK, port: 6, baudrate: 0 INFO [muorb] SLPI: sat info: disabled INFO [muorb] SLPI: rate reading: 0 B/s INFO [muorb] SLPI: Ok executing command: gps status INFO [muorb] SLPI: Sending qshell retval with timestamp 855143719, current timestamp 855143720 INFO [qshell] qshell return value timestamp: 855143719, local time: 855147735 pxh>

@Martin-Brandt Starting the RMA process is probably the best option as opening it up yourself risks damaging other components

Hi @A-Evans This is our current WiFi solution for Voxl2, or any ModalAI board with a standard USB2 4-position JST GH port. https://www.modalai.com/products/wifi-usb-cable-add-on?_fid=da227b041&_pos=1&_ss=c If you want to use our 10-pin USB3 format connector for the Alpha USB Dongle, then you can also consider using this cable: https://www.modalai.com/products/mcbl-00022 Note: We use Alpha WiFi since it's NDAA compliant You are free to use any WiFI USB dongle sourced at the UK, and make your own cables for USB2 (not too hard for proto-type builds).

@Jgaucin you can use voxl-portal to view the output of all of the image sensors on the drone. The Sentinel has forward and rearward stereo depth pre-calibrated and tested in the factory https://docs.modalai.com/voxl-portal-cameras/ this might also be useful https://docs.modalai.com/calibrate-cameras/

@Eric-Katzfey I think it probably makes a lot of sense to leave it on there and just call it out more in unboxing. We purchased a number of platforms and it only came up with 2 of them.

@Eric-Katzfey Thanks, problem solved.

@Paul-Carff Sounds like your son knows his stuff!

@Chad-Sweet Thanks. So it sounds like a full re-flash is probably the best check to re-install base software (I'm not aware of any modifications, but came onto the project after it started) and check that connection. If the issue persists with both units, kick off an RMA? Appreciate all your help.

@Neil-Huang Yes, if TBS side doesn't use 250000 then that line would need to be changed.

OMG! The protective cap on the camera You're right! Just saw it minutes before checking it here. There was no cap on the first drone; it was probably dropped or something, so I did not even feel the need to look for it. Thank you, Adrian; sorry for the trouble. Case closed!

We don't have a guide on our own, but Google should have plenty of guides and examples You'll need to add your own post processing function in this area of the code Here is Google's guide on how to postprocess tflite data https://www.tensorflow.org/lite/inference_with_metadata/lite_support

@Chad-Sweet Ok, thank you for clarifying!

The VOXL 2 Connector Specification is here: https://docs.modalai.com/voxl2-connectors/

@JustinDev23 Definitely. See here: https://www.modalai.com/pages/rma

@martygp95, Yes, powering the Sentinel with a 4S is fine. Just be sure NOT TO FLY with it. The tuning for the drone was done around the 3S and if you attempt to fly with the 4S, then there could be detrimental flight-performance. The APM on the Sentinel can handle 2S-6S [2 cell (6.4V min, 7.4V nom, 8.4V max) 6 cell ( 19.2V min, 22.2V nom, 25.2V max)]. You can find info about the APM here Hope that helps

Hello, Can you please check the vehicle's attitude (roll, pitch, yaw) when it is sitting flat, right before take-off? You can look at PX4 logs or live using GCS. If the IMU is not calibrated properly, and the vehicle thinks that it is tilted, it will try to compensate for it by tilting in opposite direction, so it can appear as if one of the motors is not able to spin faster, when in fact the flight controller could be commanding it to spin slowly. This is just one reason why this could be happening, so it is worth checking.

The tracking camera calibration worked flawlessly after using the seeker in softAP mode, disabling the dfs-server and using the correct checkboard. Stereo camera calibration was much harder, and after many attempts there was always a failure in the exstrinsics calibration. We configured /etc/modalai/extrinsics.conf following this guide https://docs.modalai.com/configure-extrinsics/ . We tried manual control with Position flight mode and it worked fine. After we tried two simple autonomous flights with mavsdk inside docker, mainly using the set_position_ned() method. http://mavsdk-python-docs.s3-website.eu-central-1.amazonaws.com/plugins/offboard.html#mavsdk.offboard.Offboard.set_position_ned First mission: arm the drone, take off 1m from the ground, move forward 1.8m and then land. This one ended pretty badly, with a drone crash. Mission was not followed at all. The environment was an office sufficiently illuminated, and the Seeker drone started from the position where it was turned on. Second mission (simpler): arm the drone, take off 1m from the ground, land. In this one the drone took off around 2.5m from the ground, and then landed with relatively high speed, hitting the ground. I don't know if it can be an issue but in both tests dfs-server was disabled.