“LusTra” comes from “ludique” in French and “sorting Trash” in English. It is a motivating and playful sorting system to educate the young public to the right gestures; thus, recycling becomes a natural reflex thanks to an interactive and rewarding gamification system. We used machine learning methods to train models on the sensor data to recognize the waste types and return the information to the child via a playful interface. This is a real challenge when we know how difficult it is to differentiate Polyethylene Terephthalate (PET) from glass, as well as to detect paper. In this first stage of the work, we mainly focused on PET detection.