Meet the Arduino Nano 33 BLE Sense

With the rise in the applications of AI and involvement from various industries, the focus on manufacturing with small form-factor is also rapidly increasing.

To boost the ecosystem companies such as Arduino, SparkFun, ST Electronics have released dedicated microcontroller boards that are beefed up with multiple capabilities fostering the growth of AI on Embedded systems.

Few such prototyping boards are:

• Arduino Nano 33 BLE Sense
• SparkFun Edge development board-Apollo3 Blue
• ST Microelectronics 32F746 Discovery kit
• Adafruit EdgeBadge

In this article, we will focus on the first one- Arduino Nano 33 BLE Sense.

This is a powerful microcontroller based on nRF52840 from Nordic Semiconductors, which is a 32-bit ARM Cortex-M4 CPU running at 64 MHz.

(Datasheet can be found at - Nano_BLE_MCU-nRF52840)

What does this board have to offer you?

• Consists of Color, Brightness, Proximity and Gesture Sensors
• 9-Axis IMU (Accelerometer, Gyroscope, Magnetometer)
• Temperature and Humidity Sensors.
• Digital Microphone.
• BLE Module.
• Embedded AI

The last capability listed above can be treated as the cherry on the top as this gives the board capability to run lightweight AI models for faster inference at the Edge (read more about- Lightweight AI Models here).

Few interesting applications using AI on uC (Arduino Nano 33 BLE Sense)

• Wake word detection- Home assistants such as Google Home, Amazon Alexa have a dedicated tiny microcontroller that runs a pretty light-weight model on Arduino Nano/SparkFun, whose task is to wake up the processor/device upon detecting the wake-word!

(source-Wake-word detection on Arduino)

• Gesture-recognition- Another interesting paradigm of applications using the inbuilt IMU is running a gesture recognition use-case on an Arduino-nano board to detect gestures for interesting applications to control appliances/home automation and empowering the disabled ones.
• Plant disease detection- Crop disease detection at an early stage will prevent agricultural losses by taking suitable countermeasures. Lightweight CNN (convolutional neural network) models can detect such anomalies and raise an alarm.

One can buy Arduino Nano 33 BLE Sense from Arduino’s official website-https://store.arduino.cc/usa/nano-33-ble-sense

To reduce the shipping cost, one other alternative for folks from India-
RS-components

We shall look into details of lightweight AI models in the next article.

Please feel free to share your feedback for improvisation and 👏 if the article was informative!