Unsupervisory Machine learning examples using ST MEMS and STM32

[kumarnaveenb]

Hi,

I am wondering does ST Micro provides any example around unsupervisory Machine learning examples to implement on board learning using ST MEMS vibration sensor and STM32.

As I can see Cartesiam provides some examples and the tools to does this work but, need to purchase their libraries /tools.

It would be great help if some one have any suggestions on above topic.

Thanks,

Naveen

[Eleon BORLINI · Accepted solution]

Hi Naaven @NK.1umar Byregowda​ ,

I've added more specific topics since I'm not an expert about unsupervised machine learning techniques.

In general, for the supervised machine learning, you can stay on the sensor side (the "edge"), or work in cloud.

Sensors' side, a technology called Machine learning core (MLC) is developed on a family of sensor such as IMUs (LSM6DSOX) or stand-alone accelerometers (. For this argument, you can refer to this application note.

There are also STM32 tools enabling neural network implementation directly on the microcontroller, and the STM32Cube.AI is the right tool that allows to convert NN developed with more conventional (python) environments (such as Tensorflow or Pandas) into C code for the STM32 architecture.

From cloud side, you can refer to the next point.

For the unsupervised machine learning, you already have noticed the NanoEdge AI solution from Cartesian, ST partner on the AI. This is indeed the solution I would have suggested you too. In alternative, you might base on cloud services such as Amazon AWS and Azure IoT. In this case, I would suggest you to have a look to a couple of sensor boards that are equipped with the connection to these services: the STEVAL-STWINKT1B (link) and the STEVAL-MKSBOX1V1.

Let me know if these solutions can help you.

-Eleon

[kumarnaveenb]

Thanks Eleon,

I will take a look at AWS and Azure side.

Naveen