The goal of this project is to ideate and build an AI toy for kids that is able to interact with the child during the game. As an additional challenge, we were asked to take inspiration from the main ML patterns, as “training”, “collaboration”, “supervising” and “evolving”. Our team decided to stick with the idea of collaboration and we designed a playful, friendly and responsive toy. Our concept is a smart evolution of the very basic and well-known balancing base game.
Interface, Technology and Interaction Modality
Winky is a balancing base with incorporated counterbalance weight inside moved by a Servomotor. An Arduino 101 with built-in Intel Curie Module detects the movements of the base and constantly recalculates the angle of rotation of the Servomotor in order to move the counterweight in a better position. Two LED rings and one LED strip form a little friendly face that gives personality to the toy and provides visual feedback to the child. The LEDs turn on and off according to the base pitch angle. The toy is completely free of buttons, since we use its ability to detect movements.
Winky is designed to look like a very simple and intuitive toy. It keeps the classic shape of a balancing base and swings happily as soon as the child touch it. No instructions and additional set up is needed to start the game. Winky’s wavy top invites to stack the wooden cylinders that we designed for the game. The smiling face is funny and cute and immediately attracts the kid’s attention. As soon as the toy starts to change its expressions, the child easily understands that they represent the different states/emotions of the toy: happy = balanced; winking on one side = increasingly slanted; confused = going to fall down or already did; sleeping = waiting patiently to play again.
Research and Development Context
Winky could become even more playful in the future, as we look forward the possibility to design different “head” tops that would fit with various types of stacking pieces and different levels of gaming challenge.