Machine Learning via Arduino

Projects General Guidance
1

Hi,

I am trying to decide what kind of kit I can make use of best. The kit will be used for anomaly detection via Tensorflow Lite. Tensorflow lists supported devices on their website. Arduino Nano 33 BLE Sense is one of them. I will be using grove sensors, hence I need a shield as well.

There is a nice Arduino Nano 33 BLE Sense shield. My only problem with this, it does not have any connectors to TX/RX, so I am not able to connect my Grove WiFi module (if there was a built-in WiFi this wouldn't be needed).

Then, I have found this MKR shield. It is the perfect shield for my case. However, even though Arduino Nano 33 BLE Sense is a member of MKR family, it does not fit on that shield (I will not be using any jumpers).

So my questions are:
1 - Is there any other Arduino modules that is fine with tensorflow lite that I can use instead of Arduino Nano 33 BLE Sense?
2 - Is there any other shield for 33 BLE Sense that enables WiFi?

2

It appears your best solution is a custom board. Some quick notes: You can use Software Serial if you can operate in 1/2 duplex. Have you looked at some of the ESP devices? The Mega has 4 serial ports if that helps.

3

The mega being a 16 bit device, cannot do TensorFLow.

4

@Idahowalker exactly. I ended up using Nano BLE Sense without WiFi. I will be checking the market if anyone produces a better shield for Nano BLE Sense.

5

ESP32's can do TensorFlow Lite. and have extra Serial ports.

6

Yes, but I could not find a good ESP32 shield where I can connect my Grove sensors.

7

This topic was automatically closed 120 days after the last reply. New replies are no longer allowed.