Arduino Nano 33 BLE Sense and Edge Impulse

Connect your Arduino Nano 33 BLE Sense with Edge Impulse

With Edge Impulse, you can easily collect sensor data, develop AI models from it, and put them back on a microcontroller. The perfect companion for this is the Arduino Nano 33 BLE Sense – not only does it have suitable sensors directly on the board, but it is also capable …

Read More
Teachable Machine Cover

Artificial intelligence for everybody – with Teachable Machine

When you think of artificial intelligence, do you have AlphaGo, Unsupervised Learning and complicated statistics in mind? Then you are right, but that does not mean that it is difficult to get into this exciting topic. Try Teachable Machine! In this tutorial you will get to know Google’s application Teachable …

Read More

Artificial intelligence is a topic that has experienced an enormous upswing in recent years – and will certainly never leave us again. Machine learning and AI models seemed to be reserved for mainframes, labs and tech companies until now. But that’s not true: you too can create artificial intelligence projects on your Arduino!

Machine learning with Edge Impulse

Of course, your Arduino doesn’t have the power to compute complex models. So it’s best to outsource this computationally intensive job. This is where Edge Impulse’s service comes in handy, and you can use it for your hobby projects for free.

Edge Impulse works with a number of microcontrollers, including the Arduino Nano 33 BLE Sense.

On Pollux Labs, you’ll learn how to connect your Arduino Nano 33 BLE Sense to Edge Impulse, for example, and use it to collect data.

Once you have the data for your artificial intelligence stored at Edge Impulse, you can calculate (have) an AI model there that you can optimize and play back to your Arduino. There you can use it for your project as usual.

Arduino Nano 33 BLE Sense – artificial intelligence made easy

As mentioned, the Arduino Nano 33 BLE Sense is perfect for your first steps into artificial intelligence.

It has a number of sensors, e.g. for temperature, pressure, colors, gestures. Two more you can currently use with Edge Impulse: The accelerometer and the microphone.

This allows you to develop, for example, an AI model that recognizes complex gestures and movements. Likewise, you can use the Arduino’s microphone to record sounds and store them directly in Edge Impulse. From this, too, you can quickly develop a neural network that recognizes, for example, a cough or a crying baby.

Develop AI models with Teachable Machine

A particularly beginner-friendly solution comes from Google: With Teachable Machine, you can develop models in no time at all. You can choose between the recognition of images, sounds and poses or gestures.

You simply train the artificial intelligence with images and sounds from your hard drive or directly with your webcam or microphone. The training itself is deliberately kept simple – however, if you wish, you can change some parameters to improve your results.

When you are done and satisfied, you can download your AI model for TensorFlow or TensorFlow Lite and use it in your projects. It couldn’t be easier.