Learning Like a Robot

In this workshop, children (4-7) become "Robot Teachers." They will use picture cards to train a robot, learning that computers need examples (data) to understand the world and sometimes they make funny mistakes.

Resources & Downloads

Download the PDF versions of this scenario or the associated game cards.

Download PDF (Document) Download PDF (Pamphlet) Download Game Cards & Award Certificate (.zip)

Working Editable Documents

These editable source files are provided as working documents for teachers.

Teacher Guide Source (.docx)

Target Group: 4 - 7 y.o.

Activity Duration: 3–4 sessions (30–45 min each)

Key Learning Goals:

Machine Learning: Understand that computers learn from examples, just like children do.
Data & Training: Learn that computers need humans to show them pictures (data) to learn.
Error Handling: Notice that robots can make funny mistakes if they don't have enough examples.
Collaboration: Work together to "teach" the robot by sorting cards.

Learning Outcomes

By the end of the project, students will be able to:

KNOWLEDGE & UNDERSTANDING:

Know that machines can "look" and "learn" from pictures[cite: 325].
Understand that computers need humans to show them examples[cite: 325].
Notice that sometimes machines get things wrong[cite: 325].

SKILLS & ABILITIES:

Sort and group images (Fruit vs. Veg, Animals vs. Birds)[cite: 325].
Describe what they see in pictures[cite: 325].
Notice patterns or differences between objects (e.g., colours, shapes)[cite: 325].

ATTITUDES & VALUES:

Feel curious about how robots and computers work[cite: 325].
Learn to share ideas and listen to others in a group[cite: 325].
Recognise that people and robots can work together[cite: 325].

European Dimension / Erasmus+ Connection

Digital Inclusion: Making complex concepts like Machine Learning accessible to very young children.
Scientific Collaboration: Simulating real-world research where diverse data leads to better results[cite: 325].
Critical Thinking: Early introduction to the idea that technology is not perfect and relies on human input.

1. Resources and Tools

Required Materials:

Printable Cards: Fruits, Vegetables, Animals, Birds, Emotions (included in download)[cite: 325].
Sorting Mats: Hula hoops or labelled tables (e.g., "Fruit" vs "Vegetable")[cite: 325].
Props: A "Robot" puppet or costume for the teacher[cite: 325].
Storybook: "Boy + Bot" or Sesame Street clips[cite: 325].

Useful Links:

Classroom Support Materials:

Padlet can be used for sharing class reflections, and eTwinning or TwinSpace can be used if partner classes are exchanging simple outputs.

2. Working Methods

Gamification: Using sorting games to mimic "training" an AI[cite: 325].
Role-Play: Teacher acts as the "Robot" making guesses based on student input[cite: 325].
Storytelling: Using stories to reinforce that robots need human help[cite: 325].
Hands-on Learning: Physical sorting of cards into categories[cite: 325].

Activity Overview

Phase	Activity	Description
1. Intro	Train the Robot (Sorting)	Students work in teams to sort cards (Fruits vs. Vegetables) into hoops. This represents "feeding data" to the robot.
2. Testing	Robot Guesses	The teacher (as Robot) picks a card and guesses what it is. Students say "Correct" or "Error" (if the robot confuses a tomato for a fruit).
3. Advanced	Human vs. Robot	Sorting tasks into "Robot Job" (calculating) vs "Human Job" (hugging, feeling). Using emotion cards.
4. Reflection	I Helped the Robot	Drawing a picture of how they taught the robot something new.

3. Introduction and Motivation

Activity 1: Train the Robot (Data Sorting)

Goal: Teach students that computers learn from examples.

The Hook: "I have a robot friend, but he doesn't know what a strawberry is! Can we teach him?"

Setup: Place hoops on the floor labelled "Fruit" and "Vegetable".
Action: Students take turns placing cards into the correct hoop.
Explanation: "Good job! You are giving the robot examples so it can learn."

4. Research and Learning

Activity 2: Robot Guesses (Testing the Model)

Goal: Show that robots only know what we teach them.

Role-Play: The teacher puts on a robot voice/hat.
The Test: The robot picks up a card (e.g., a Carrot). Robot says: "This is... a Fruit!"
Correction: The children shout "No!" and explain why (e.g., "It's a vegetable because it grows in the ground").
Learning Moment: Explain that the robot made a mistake because it needs more training.

5. Creative Application

Activity 3: Human vs. Robot (Emotions)

Goal: Distinguish between computational tasks and emotional intelligence.

Sorting Game: Use "Action Cards" (e.g., Adding numbers, Giving a hug, Knowing the colour red).
The Question: "Can a robot do this, or do we need a person?"
Key Takeaway: Robots are good at math and patterns, but humans are best at feelings and kindness[cite: 325].

6. Reflection and Evaluation

Activity 4: Creative Reflection

Task: "I Helped the Robot!" Drawing.

Students draw a picture of themselves teaching the robot.
Captions: Teacher helps write captions like "I showed the robot a green apple" or "I told it the dog was not a cat"[cite: 325].
Display: Hang pictures on a "Robot Helpers" wall.

Co-founded by the Erasmus+ Programme of the European Union • 2024-1-HR01-KA220-SCH-000254267
This work is licensed under a Creative Commons Attribution 4.0 International License.