Issue with NanoEdge AI Model Training on STWINBX1 (STEVAL-STWBXCS1A)

[Mahdouch]

Hello ST Community,

I'm working on a Predictive Maintenance project using FP-SNS-DATALOG2 to capture real-time sensor data from my STWINBX1 (STEVAL-STWBXCS1A). Here’s my workflow so far:

1. Data Acquisition: Captured sensor data using HSDatalog_GUI.
2. Format Conversion: Converted data to NanoEdge CSV format using " hsdatalog_to_nanoedge.py ".
3. NanoEdge Processing: Uploaded the converted data to NanoEdge AI Studio and successfully downloaded a benchmark.
4. Model Training Issue:
   • I connected my board via USB-C and STLINK-V3MINI (otherwise, my PC and NanoEdge wouldn’t detect it).
   • My STLINK is properly recognized on COM3.
   • However, when I try to train the model, NanoEdge AI Studio does not recognize the board—it detects nothing.

Has anyone encountered this issue before? Could it be a driver issue, a missing configuration, or something else? Any help would be greatly appreciated!

Thanks in advance!

@Andrew Neil 

@SimonePradolini 

[ahsrabrifat]

Open STM32CubeIDE, select ST-LINK, and check if your STWINBX1 is detected. Open NanoEdge AI Studio and check if it is set to detect devices on the right COM port. Try switching to a different USB port on your PC. Restart NanoEdge AI Studio after plugging in the board.

[Mahdouch]

Thanks for your response! I’ve already tried that method.

• I’m working on a laptop with only two USB ports, so I’ve tested switching between USB-C and STLINK-V3MINI multiple times.
• No matter which port I use, COM3 always appears in Device Manager, but the issue in NanoEdge AI Studio remains the same—it doesn’t detect the board.
• In STM32CubeIDE, my board is detected through STLINK, and I can debug successfully.
• However, when I open PuTTY (COM3, 115200 baud rate), nothing appears—it's like the board isn’t communicating at all.
• Interestingly, when I use STM32CubeProgrammer, I can successfully flash firmware via STLINK without any issues.

It seems like the board is recognized for flashing/debugging but not for serial communication or NanoEdge AI Studio. Could this be a firmware or driver issue? Any ideas on what else to check?

[SimonePradolini]

Hello @Mahdouch 

FP-SNS-DATALOG2 and NanoEdge AI Studio aren’t working in the same way:

• FP-SNS-DATALOG2 streams live data via USB only. Live data streaming is managed by the hsdatalog_GUI aforementioned.
• NanoEdge AI Studio supports live data streaming also, but via COM port. The tool itself offers you a specific firmware that allows data streaming via COM port through ST-LINK.

Notice that the two firmwares are different and not compatible with each other: by using FP-SNS-DATALOG2 firmware, you can't live logging data with NanoEdge AI Studio and viceversa.

Step 1-3 you described seems ok. Which firmware are you using for your step 4?

Best regards,

Simone

[Mahdouch]

Hello @SimonePradolini 

I'm currently using the "DATALOG2_release.bin" found in "FP-SNS-DATALOG2-V2.3.0\STM32CubeFunctionPack_DATALOG2_V2.3.0\Projects\STM32U585AI-STWIN.box\Applications\DATALOG2\Binary"

should i use another firmware for NanoEdge AI studio  detection (so i can't train my model using my real time data coming from the board ?) , if so can you provide me with that firmware so i can flash it on my board .

Thank you Very much.

[Julian E. · Accepted solution]

Hello @Mahdouch,

I would suggest you to follow this tutorial: 

https://wiki.st.com/stm32mcu/wiki/AI:How_to_Build_an_Anomaly_Detection_Project_for_Predictive_Maintenance_with_NanoEdge_AI_Studio 

The board used in this case is the STEVAL PROTEUS, but you can use yours as it is also available in the NanoEdge Datalogger generator.

This tuto goes from logging data to deploying the model on the board, so you will find everything needed.

Have a good day,

Julian

[Mahdouch]

THANKS , @SimonePradolini 

Have a good day !