@Nikos-Mavr You're welcome! And can you please start a new forum post for this new issue? Thanks!

@Cliff-Wong said in tarling - Path planning (blue line) is erratic and drone moves to wrong locations:

/etc/modalai/vio_cams.conf

Hi there, thanks for the detailed suggestions!

I have gone through all your steps to discard physical and configuration issues, but it seems I'll need to upgrade. Here is the checklist of what I tested:

Camera orientation: I checked the front camera in voxl-portal and it is completely upright (tops are aligned, ceiling is up, floor is down).

Physical mounts: The VOXL 2 unit is mounted correctly from the factory (USB-C port facing the ground, WiFi dongle on the top side).

IMU Calibration: I ran a full 6DOF voxl-calibrate-imu on a flat surface. It completed successfully, and voxl-inspect-imu shows perfect readings now (X and Y at ~0.00, and Z gravity at exactly -9.80).

VIO config: I checked /etc/modalai/vio_cams.conf. tracking_down is disabled/not present, and it is only using the front camera as expected. Here is my exact configuration:

{
"cams": [{
"enable": true,
"name": "tracking",
"pipe_for_preview": "tracking",
"pipe_for_tracking": "tracking",
"is_occluded_on_ground": false,
"imu": "imu_apps",
"cal_file": "opencv_tracking_intrinsics.yml"
}]
}

Since all physical and configuration checks passed perfectly, I am going to follow your advice and proceed to flash SDK 1.6.3 to take advantage of the advanced config installer and move to OpenVINS.

Thanks again for the help, I'll keep you posted on how the 1.6.3 update goes!

Hi @austin-c ,

You can find the latest M0138 firmware with regen completely disabled here

Also, I added the esc config that i've been using for testing here

For step tests, i have been using pretty aggressive step command:

./voxl-esc-spin-step.py --id 2 --rpm 2000 --step-amplitude 4000 --step-frequency 2.0 --timeout 2.0 --cmd-rate 2000 --enable-plot 1

In order to make the ESC not de-sync (due to long demag time of the motor), the following params were adjusted:

max_rpm_delta set to 4000 (you can go lower), to cap the maximum rpm transitions. Although in my test, the step is 4000, so this had no effect timing_advance: 45 -- this is pretty high commutation advance to trigger commutation earlier so that there is more time for back-emf sensing -- high advance wiil reduce motor efficiency and will cause the motor to heat up a bit more sense_advance: 30 -- delay back-emf sensing so that we don't sample back-emf during the motor coil demagnetization more details : https://gitlab.com/voxl-public/voxl-sdk/utilities/voxl-esc/-/blob/dev/voxl-esc-tools/doc/low_kv_motor_tuning.md

I had another thought.. What if you use a small battery to absorb the voltage spikes? Basically hook up a 6-7S battery with a beefy Schottyky diode (10A) set up with Cathode (negative) pointing to the battery.

you will have to make sure the power supply voltage never exceeds battery voltage + diode drop. you will also need to make sure that you don't over-charge the battery with repeated spike absorption.. It is possible to have some sort of self-discharge circuit to slowly discharge the battery to a certain voltage (using a Zener diode + resistor or something similar).

Anyway, just an idea..

Try out the no-regen firmware..

Alex

@Myles-Levine , if you are masking out 90% of down-facing camera, then it is probably useless any movement of the features will get it out of the unmasked region very easily and feature will be dropped.. It may hurt VIO as there may be features going in and out, only trackable for a few frames (just adds to the complexity).

You should try to move the down-facing camera to free up it's FOV or even have it facing slightly angled to the back of the drone..

Alex

@Dan-Jennings I'm also guessing that the driver may not be detecting the hardware so is not able to get the data for the battery_status. Have you tried swapping some other hardware to see if the problem is associated with a particular unit?

@emarcphera @Eric-Katzfey thank you so much after upgrading to voxl-mainline im getting GPS readingScreenshot from 2026-02-28 01-45-10.png

@RyanH ,

Sometimes there is small pieces of random magnetic debris that gets stuck in the motor.

First, you could try using compressed air to clear out the motor.

Additionally, It could help to spin the motor up and down using voxl-esc command line tools a few times to clear it out (without prop is best, just for safety). Spinning the motor without prop a bunch of times could get the debris dislodged and actually broken up by the moving motor parts, so that the debris is no longer large enough to cause a problem. Reversing motor spin using voxl-esc test tools can help dislodge the debris (you can commend negative power/rpm).

You can find useful info regarding motor testing here : https://docs.modalai.com/voxl-escs/faq/

Alex

@bendraper
Hello,
I hope this quick clarification helps for warnings:

CMake Policy warnings: OV internally (external dir in the repo) still uses boost and references an older cmake version -- it's a harmless warning for devs (as you are building from scratch)

Yaml-cpp lib warnings: This is a third party dependency that gets compiled from scratch during the make/build process. We use C++ 17 but the library is based on an older C++ version; nevertheless, it compiles fine on C++ 17 with minor warnings. -- it's a harmless warning

TrackOCL warning: ModalFlow is a core library written in C++ 17 that allows function call usage as done in TrackOCL, i.e., the narrowing conversion is harmless in this case.

That being said, we are actively working on the dev branch, so all flight--impacting warnings are hopefully caught early. We currently have no flight-impacting warnings/bugs identified, but if you found one warning/bug that caught your attention, please feel free to share.

All the best,
ZBFT

@restore There are a few examples of code that listens for Mavlink messages in our SDK. For example, in voxl-mpa-tools, take a look at the source code for voxl-inspect-mavlink.c. https://gitlab.com/voxl-public/voxl-sdk/utilities/voxl-mpa-tools/-/blob/master/tools/voxl-inspect-mavlink.c?ref_type=heads

Probably would be a good idea to run voxl-inspect-mavlink on the mavlink_onboard and mavlink_to_gcs pipes to see which one has the desired Mavlink message in it. Then create your own application that listens for the Mavlink message on that pipe.

@Alex-Kushleyev The Alfa Network AWU2036

@qubotics-admin First you need to determine the version of PX4 running. With ssh or adb you can run the following at the command line: px4-ver all. You will see PX4 version and Vendor version. So, for example, for px4 version 1.14.0 and vendor version 8.134.3 then you would checkout the tag v1.14.0-8.134.3-dev from the repo to see that commit.

@ndwe Whether you use a VOXL or Jetson as a companion computer shouldn't make any difference unless you want to use the VOXL SDK. The version you run is totally up to you. If our v1.14.0 based version works for you then use it. We do plan to transition to v1.17.x in the near future. I don't think the issues you are seeing are related to something specific in the flight core v2 hardware setup. I would think that if you swapped it out for some other brand you would see the same issue. I would ask on the PX4 forums to see if others have experienced the same issues.

Hi @Alex-Kushleyev, sorry for the late reaction, it took some time before i could update to the new SDK (1.6). the system now works normally with the resolution 2704x1540, and downscaled to 1920x1080 for the MISP pipeline. Thanks for the help!

@Piyush-Singh , can you please provide more details

how exactly all cameras are connected to VOXL2 mini (including interposers, flexes), etc. A diagram would help a lot which kernel variant are you running on your voxl2 mini (2.0.0 or 2.0.2).

Are all the other cameras working if you disable the TOF sensor in the camera server config?

Alex

@cbay voxl-fpv-px4 is based on voxl-px4 but has some custom tuning code / parameters for our fpv drones that isn't in voxl-px4. So, it's kind of a superset. But there isn't anything related to low level i2c drivers so I have no idea how that could be causing this issue.

@Jetson-Nano , since we currently have no plans to do the MIPI integration of this sensor, you always have the option to use USB connection and voxl-uvc-server to publish images via MPA. You would need to check if the sensor supports the standard UVC interface and try it out...

Alex

@cbay , yes this should work.

Whenever you add cameras of the same type, you should always check for possible I2C (CCI) conflicts. If you look here https://docs.modalai.com/M0188/#image-sensor-interfaces , the sensor IDs 0,2,3,6 (which is what you use) use unique CCI buses 0-3, so that is all good.

If you used ID 1, it would conflict with with another AR0144 on ID6 (both use CCI1)

One last thing to check is the sensor sync.

if you see fsin in the configuration of the sensor (and name of the sensormodule in /usr/lib/camera/), it means it is set up as a sync slave currently you have the IDs 0 and 6 set up as sync slaves, but ID 2 and 3 as free-running

There are two issues enabling sync'ed drivers for AR0144 in ID 2 and 3:

the driver is provided only for ID2 in : /usr/share/modalai/chi-cdk/ar0144-fsin/ - this is easy to fix ID1 and ID3 on M0188 do not have the sync line connected to the camera connector this is easy to fix by installing a 0-ohm resistor, which is DNI by default.

Depending on your requirements for VIO (which VIO are you using?), you may or may not need the cameras synchronized.

Please let me know your requirements and we can go from there..

Alex

@Jskim , you should check dmesg to see maybe there are some warnings. Perhaps the device is not de-initialized properly (how are you powering it off?) and next time the OS allocates a new Device ID.

Either way, using udev rule should help resolve the issue, however make sure there is no resource leak as a result of the device on/off. Also, if you keep turning the device on/off, does the device ID increment indefinitely?

Alex

@Alex-Kushleyev Thank you for the hint. We have fixed the issue by changing the resolution to 2704x1540 in the voxl-camera-server.conf. Now, the actual DFOV is 150°.

Hi @Vinny

Thank you for your reply! That's good news:)
We are looking forward to it!