Available options to load a new model in LSM6DSOX (of SensorTile.box)

[martinj3456]

Hello all,

I have bought a SensorTile.box which has LSM6DSOX, an IMU with Machine Learning Core. I am seeing some decision tree-based examples/implementation of LSM6DSOX.

If I train a new model for my application, My query is how to load that model into LSM6DSOX of SensorTile.box. General examples of LSM6DSOX says I need to insert the LSM6DSOX (DIL 24 socket) in an ST MEMS adapters (STEVAL-MKI109V3) and use Unico tool. How to do that for SensorTile.box?

Another query is whether any other low-cost option available to load a new decision tree model into the LSM6DSOX chip without buying that STEVAL-MKI109V3.

Thanks in advance.

[Eleon BORLINI]

​Hi @martinj3456​ , I immediately tell you that it is not so simple to implement the MLC with the DSOX of the SensorTile.box. We are developing the related app function, selectable from the STBLESensor app, but I believe they'll be present from the next release on. By now, the "simplest" method, as you are already suggesting, is to start from the STEVAL-MKI109V3 + Unico (or the X-NUCLEO-IKS01A2), manage the acquisition / decision tree, generate the .c file and implementing it in the SensorTile.box FW, maybe starting from the FP-AI-SENSING1 libraries (i.e. the neural networks for real-time human activity recognition (HAR) and acoustic scene classification (ASC) applications). Not so simple by now, but will be in the near future. Regards

[AKato.1]

Hi!

I have been fighting with the same issue recently. I think, I have found a solution:

• Start a new .ioc project STM32CudeIDE
• Add additional SW -> AI Core + Template Application
• Add additional SW -> MEMS - LSM6DSOX (SPI) + Custom Motion Sensor
• Configure MEMS - LSM6DSOX: PE12 + SPI1

Than you can use function some custom_motion_sensors.h to collect data from the sensors.

As a next step, I will modify MX_X_CUBE_AI_Process() and feed my NN with real sensor data.

Best regards,

Attila