STEM and Machine Learning: Creative Projects in Elementary School

​[[{“value”:”

CC-BY, provided by author

In my view, applying STEM (Science, Technology, Engineering and Mathematics) at elementary level is an incredibly exciting experience. The children’s curiosity and investigative spirit turn every lesson into a small journey of discovery. Using accessible and fun tools such as Google’s free Teachable Machine greatly simplifies the process of illustrating the basics of artificial intelligence and machine learning through concrete examples. Whenever we work together to find answers to the question, ‘What can we do in real life?’, the excitement on the children’s faces motivates me every time.

Previously, we focused on the issue of our schoolyard flowers falling ill and devised a coding and machine learning solution to tackle this problem. Once we had combined observations, images and labels for diseased and healthy flowers, a fun AI application emerged. However, flowers aren’t our only focus! When students ask questions such as ‘How can we reduce traffic accidents?’ or ‘How can we identify healthy foods?’, we apply the same approach to different projects.

For example, in a project about traffic, we categorise images of streets or found online as either ‘safe traffic conditions’ or ‘risky traffic conditions’. Using Teachable Machine, students can train the model using specific photographs. Then, when a new photo or short video is uploaded, our model tells us whether it is a risky or safe situation, guiding us in real time. Thus, the students grasp how machine learning works and start generating ideas for making traffic safer.

 

CC-BY, provided by author

 

Projects related to health and food are equally popular. For instance, we have designed a model that can distinguish between different food items captured on store shelves, classifying them as either ‘fresh fruit’ or ‘starting to go bad’. After uploading and training these images in Teachable Machine, the children hold up the fruit they have brought to the next class and ask, “Is this still fresh?” When the model’s predictions are incorrect, we ask questions such as ‘Why was it wrong?’ and ‘What data are we missing?’. In doing so, they discover that mistakes are a natural part of science and the process of trial and error.

What I love most about these projects is that every student has an important role to play. Some gather pictures, some handle labelling and others assemble the code blocks or add sound effects to the project. In fact, some students suggest things like adding an alarm sound so that when the flower is sick, it says, “Help me!” This way, children become designers and developers as well as users of technology.

 

CC-BY, provided by author

 

This lively learning environment has truly transformed lesson time at our school in Ankara, Türkiye. In science class, we study the structure of plants; in technology class, we develop machine learning models; and in maths class, we calculate how much water a plant needs or analyse traffic statistics. Children never need to ask, “What’s the point of this lesson?” because we always focus on solving a real problem. This way, they experience firsthand how what they learn can be applied in everyday life.

 

CC-BY, provided by author

 

If you’d like to carry out similar projects at the elementary level, I suggest starting with a simple subject. For instance, you could create a basic machine learning model that differentiates between fresh and spoiled food in the school cafeteria, then move on to a traffic safety project, and later address flower diseases. Children maintain an unceasing curiosity while they get an early introduction to future technologies—boosting their creativity along the way.

Let’s not forget that the main goal of STEM is to equip students with analytical thinking and problem-solving skills. Thanks to user-friendly tools like Teachable Machine, children break the notion of “Artificial intelligence is too complicated” and confidently explore this new realm. Seeing the excitement on students’ faces when they exclaim, “I can do this, too!” after each project is the best reward for all our efforts. Each lesson becomes a small yet powerful step toward the major projects of the future.

 

 

 

About the author

Hüseyin Sihat is a STEM educator and Scientix Ambassador from Türkiye, with expertise in AI, coding and teacher training. He has been involved in STEM education since 2015 and actively contributes to national and international projects, including Erasmus+ and TÜBİTAK. He currently teaches at the Sincan Science and Art Centre (BİLSEM), working with gifted students on artificial intelligence, robotics and digital citizenship.

The post STEM and Machine Learning: Creative Projects in Elementary School appeared first on Scientix blog.

“}]]