Our literacy learning approach is centered on child-driven interactions with an open-ended, expressive app, where children spell words to make scenes and stories. In our previous studies, we saw how children created stories via a kind of back-and-forth interaction between them and the app: some ideas originated from them, while some were inspired by the medium. We think that this dynamics can be reinforced by adding co-creation capabilities to the app. This might lead to better engagement in creative activities and alleviate the "fear of the blank canvas", which is one of the barriers creators of all kinds have to face. Robotic companions seem to be ideal in the role of a co-creative partner.

To that end, we added an "idea button" to the interface of the app. Pressing on it prompts the robot (on this picture, a virtual one) to give suggestions on what the child can make. If there is only an empty canvas, it will give a starter idea, for instance: "I think airports are very cool! Let's make a picture of an airport?" If there are already objects on the scene, the robot will suggest something that is contextually fitting. For instance, tapping the idea button in the situation shown below caused the robot to give a prompt: "Who steers the plane? Let's add this person to the picture!"