After manual Restart IMU Sensors crash

Dear Stereolabs Team

We have since achieved a more stable startup program for the Fusion Stuart.

We wanted to restart the fusion after performing some PC maintenance today, but we ran into an issue.

We wonder whether this is a malfunctioning camera or if it’s completely random.

The log we obtained in our terminal is as follows:

Failed to open  C:\Users\autstart-wardobe\Desktop\ZED_Unity_Livelink_Fusion.json.yml
[FusionSender] Waiting for ZED camera SN 31202332 to be ready...
[2026-01-21 10:04:53 UTC][ZED][WARNING] Failed to open C:\Users\autstart-wardobe\Desktop\ZED_Unity_Livelink_Fusion.json.yml
[2026-01-21 10:04:53 UTC][ZED][INFO] Logging level INFO
[2026-01-21 10:04:55 UTC][ZED][INFO] [Init]  Depth mode: ULTRA
[2026-01-21 10:04:56 UTC][ZED][INFO] [Init]  Camera successfully opened.
[2026-01-21 10:04:56 UTC][ZED][INFO] [Init]  Camera FW version: 1523
[2026-01-21 10:04:56 UTC][ZED][INFO] [Init]  Video mode: HD720@30
[2026-01-21 10:04:56 UTC][ZED][INFO] [Init]  Serial Number: S/N 31202332
[FusionSender] camera SN 31202332 is ready.
Failed to open  C:\Users\autstart-wardobe\Desktop\ZED_Unity_Livelink_Fusion.json.yml
[FusionSender] Waiting for ZED camera SN 34413725 to be ready...
[2026-01-21 10:04:57 UTC][ZED][WARNING] Failed to open C:\Users\autstart-wardobe\Desktop\ZED_Unity_Livelink_Fusion.json.yml
[2026-01-21 10:04:57 UTC][ZED][INFO] Logging level INFO
[2026-01-21 10:04:58 UTC][ZED][INFO] [Init]  Depth mode: ULTRA
[2026-01-21 10:04:59 UTC][ZED][WARNING] [ZED] Sensors could not be initialized. Some modules (AI-based) might not work correctly.
[2026-01-21 10:04:59 UTC][ZED][INFO] [Init]  Camera successfully opened.
[2026-01-21 10:04:59 UTC][ZED][INFO] [Init]  Camera FW version: 1523
[2026-01-21 10:04:59 UTC][ZED][INFO] [Init]  Video mode: HD720@30
[2026-01-21 10:04:59 UTC][ZED][INFO] [Init]  Serial Number: S/N 34413725
Error: MOTION SENSORS REQUIRED
[2026-01-21 10:05:00 UTC][ZED][WARNING] MOTION SENSORS REQUIRED in enum sl::ERROR_CODE __cdecl sl::Camera::enableBodyTracking(struct sl::BodyTrackingParameters)
Error: INVALID FUNCTION CALL
[2026-01-21 10:05:10 UTC][ZED][INFO] Logging level INFO
[2026-01-21 10:05:10 UTC][ZED][WARNING] INVALID FUNCTION CALL in enum sl::ERROR_CODE __cdecl sl::Camera::open(struct sl::InitParameters)
Error: INVALID FUNCTION CALL
[2026-01-21 10:05:20 UTC][ZED][INFO] Logging level INFO
[2026-01-21 10:05:20 UTC][ZED][WARNING] INVALID FUNCTION CALL in enum sl::ERROR_CODE __cdecl sl::Camera::open(struct sl::InitParameters)
Error: INVALID FUNCTION CALL
[2026-01-21 10:05:30 UTC][ZED][INFO] Logging level INFO
[2026-01-21 10:05:30 UTC][ZED][WARNING] INVALID FUNCTION CALL in enum sl::ERROR_CODE __cdecl sl::Camera::open(struct sl::InitParameters)
Error: INVALID FUNCTION CALL
[2026-01-21 10:05:40 UTC][ZED][INFO] Logging level INFO
[2026-01-21 10:05:40 UTC][ZED][WARNING] INVALID FUNCTION CALL in enum sl::ERROR_CODE __cdecl sl::Camera::open(struct sl::InitParameters)
Error: INVALID FUNCTION CALL
[2026-01-21 10:05:50 UTC][ZED][INFO] Logging level INFO
[2026-01-21 10:05:50 UTC][ZED][WARNING] INVALID FUNCTION CALL in enum sl::ERROR_CODE __cdecl sl::Camera::open(struct sl::InitParameters)

After we started the ZED SDK Sensor Lab we saw that the one camera with the error is showing a very weird behavior, i made a video of it:

After starting the ZED Explorer also showed the Firmware version is -1 on that camera.

It took us several time to reset the USB devices until suddently it worked again.
After a while it also showed a firmware version again.

Is it damaged maybe? Any ideas?

Best Regards

Pascal

Hi @pachermann
This problem is normally caused by a bad initialization of the MCU that handles the data acquisition of the sensors.

Please open the camera with ZED Explorer and verify the MCU firmware version.
If it reports 777, please upgrade it to 778.

Please wait a few minutes for the process to complete, even if the ZED Explorer tool may look frozen.

1 Like

Thank you @Myzhar we already tried to plan a date to go to the museum to check and upgrade the firmwares.

Do we have to re train the AI-models on the Camera after this?

Best regards

Pascal

@Myzhar

We will try patching the camera tomorrow, can i also ask you, do we need to calibrate the cameras in any way before we use them for the fusion?

we use the unity fusion setup with 2 cams, and all we do is the calibration of the Room Setup with fusion 360 else from that we never did anything.

No. This is only recommended if your camera received a shock that could have modified the optics configuration, or if you are using it in a non-standard way; e.g., underwater, with additional lenses, behind non-flat glass, etc.

No, this is not required.

Thank you very much!

Something else i wanted to ask you about the Cameras, we noticed since winter the light is not constant in the muesum, sometime we have glares/blooms and eventually even lensflares in both the stereocams.
In summer the tracking was super nice and accurate.
Since winter then the body tracking in fusion is not working specially the legs are terrible.

is there anything else, then, covering up the whole experience to stabilize this?
We tried cover the white back with a towel and i went much better.

this image show the cams and my proposal to the museum where they need to cover the cams view.

ON a sunny day its even worst, the conrete floor is all white burnt out, also the backwall is making shadows and its super white bloming.

Are you using cameras with or without the polarizing filter?

We have no polarization filters.

Normally, it’s not required indoors. But if the area is subject to direct sunlight, it can improve the behaviors.

One thing that you can do now is to use the Region of Interest setting for the auto exposure and select a zone of the image to be used as a reference for tuning the exposure.
This will allow you to not have dark important areas when the sunlight hits zones that are not important.

More info here: AEC_AGC_ROI

Examples: C++/Python