How to use source code from Nordic SDK for Mesh in the Arduino IDE for the Nano 33 BLE?

[Aduka_27]

Hi guys,

As long as I know the Arduino Nano BLE 33 is built upon the nRF52840 microcontroller, so I am wondering now is it possible to use the source code from Nordic SDK for Mesh in the Arduino IDE? I have experience working with nRF52840 DK on which I enabled the Bluetooth Mesh and created an application that enables the communication between two BLE nodes. Now the new project is divided into 2 main parts: the tinyML part that I have already enabled on my Nano BLE 33 board by using the TensorFlow library and existing examples, and the communication part for which I am planning to use the same Mesh implementation that I considered with the nRF52840 board. Is it possible to use the code written in Segger for Mesh in Arduino IDE? Also, can TensorFlow Lite be ported to run in Segger or is there a Segger compatible library that achieves the same objective in case I cannot use the Mesh implementation from Segger (Nordic SDK for Mesh)? In the end, if Segger code cannot be ported to Arduino IDE, is there any similar implementation that can be run in Arduino IDE and that can enable the Mesh communication? Any kind of help is welcome!

[Edvin]

Hello,

I am not familiar with how Arduino IDE works "under the hood", but this is not something that we have done/are going to do. Perhaps you can check with Arduino whether they have any mesh support in their stacks.

Best regards,

Edvin

[Aduka_27]

Hello,

Aha okay, no problem, I understand. But is there maybe any library for Nordic boards that can enable the same features that I have on my Arduino Nano BLE 33? For example, the library that can support the person detection model from TensorFlow Lite?

[Edvin]

Hello,

Sorry for the late reply. We don't supply anything that is working with the Arduino IDE, but you can check with arduino if they have anything for TensorFlow. Does the Arduino Nano BLE 33 have a debugger, or is it using a serial bootloader? If it has a debugger, I guess you can use whatever you like to program it (so you can compile in NCS/nRF5 SDK and program it manually). If not, I guess you need to conform with the Arduino bootloader, which means you probably need to compile the application in the Arduino IDE. 

BR,

Edvin

[Aduka_27]

Hi Edvin,

No worries, I had to finish some other things before I was able again to work on my Arduino board. As long as I know there is only support for Arduino IDE and the board does not come with an external debug interface connector but there are some required connections that can be used together with JLink Plus Compact. I am not sure if it is possible to use the Arduino Nano 33 BLE in combination with Segger using this approach. Do you maybe have some information about this implementation? 

Related to TensorFlow Lite part, there is already the library created specifically to Arduino boards that can enable some TinyML implementation on the board. I have already tested it and there are no issues at all. But now my concern is if this library can be used together with Nordic code related to Mesh implementation. I do not know if you are familiar with this part as well. 

Again any kind of help is welcome in this situation. 

Cheers,

Adnan

[Edvin]

Aduka_27 said:

Do you maybe have some information about this implementation? 

I don't know. Neither of these devices (apart from the nRF chip itself) is made by us. How they have laid out the board (arduino nano 33 BLE) I don't know. And I don't know what possibilities your have to attach an external debugger.

Aduka_27 said:

But now my concern is if this library can be used together with Nordic code related to Mesh implementation. I do not know if you are familiar with this part as well. 

I don't know. You need to ask Arduino if they  have any mesh capabilities.