AI-Driven Robotic Flower

Summary

Students build a smile-recognition model in Python and use it to control a robotic arm they assemble and design.

K-12 Learning Goals

Basics of ML + human-centered AI; training/testing models and using AI to control physical systems.

Undergraduates Learning

Prototype build + ongoing curriculum development/support.

Overview:

The Robotic Flower is an interactive installation that blooms when you smile — blending computer vision, TensorFlow, machine learning, and robotics into a real-time, expressive experience. Built using a Raspberry Pi, a custom-trained CNN (Convolutional Neural Network) for facial expression recognition. We use an off-the-shelf robotic arm kit that is decorated into a flower. The flower reacts to human emotion in a playful, artistic way.

Activity Details:

Focus: Build a smile-detection model and use it to control a robotic flower

Output: Trained model + working flower that blooms/wilts in real time

Skills: Python, basic ML (train/test), computer vision, Raspberry Pi, robotics + iteration

Setting: High-school classroom or lunch-time program (hands-on workshop)

Duration: ~60–90 minutes (or 2 sessions for build + train/test)