Discover our
resources

‍

This first activity invites you to design and train your own autonomous robot using LEGO Spike and AlphAI software. It's a great way to discover artificial intelligence in a way that's practical, fun... and accessible to everyone!

🧠 Step 1: Imagine and build your robot

Let your imagination run wild! Build a LEGO Spike robot the way you want it:

• Choose your sensors (ultrasonic, color, button...)
• Decide what actions it should perform (move forward, turn, make a sound, etc.).

Once your robot has been assembled, connect it to the AlphAI software as explained in the previous section.

🎛️ Step 2: Configure the sensors

In AlphAI, click on the "Sensors " tab in the blue vertical navigation bar on the left.

Select the sensors actually present on your robot to enable the software to take them into account during training.

⚙️ Step 3: Define possible actions

Now go to the "Actions" tab, still in the left-hand vertical menu.

Choose the actions your robot will be able to perform, depending on the situations it encounters.

📚 Step 4: Train your robot (Supervised learning)

Activate "Learning" mode.

When the sensors are stimulated (or not), assign a corresponding action to each situation.

So, you label the data, i.e. you show AlphAI how to react in each case. The software then builds an artificial intelligence model, based here onsupervised learning.

🧩 The neural network used in this activity is deliberately simple: one input layer (sensors), one output layer (actions), and no hidden layers.

You can observe the evolution of learning by clicking on the "Visualization" tab to see either real-time data or the structure of the neural network.

🚀 Step 5: Let the AI take over

When training is complete, switch to stand-alone mode.

Your robot then acts independently, guided only by the artificial intelligence you've trained.

🎯 Teaching objective

This activity highlights the power of AI in educational robotics:

→ In just a few minutes, you bring to life a robot capable of reacting intelligently to its environment.

→ Where it might take hundreds of lines of code, AI makes it possible to obtain relevant behavior by simple demonstration.

That's the power of machine learning: it allows you toautomate intelligence without technical complexity, and that's exactly what you'll experience here.

‍