CoWriter

Learning to write with a robot

Two children teaching Nao

The CoWriter Project aims at exploring how a robot can help children with the acquisition of handwriting, with an original approach: the children are the teachers who help the robot to better write! This paradigm, known as learning by teaching, has several powerful effects: it boosts the children’ self-esteem (which is especially important for children with handwriting difficulties), it get them to practise hand-wrtiing without even noticing, and engage them into a particular interaction with the robot called the Protégé effect: because they unconsciously feel that they are somehow responsible if the robot does not succeed in improving its writing skills, they commit to the interaction, and make particular efforts to figure out what is difficult for the robot, thus developing their metacognitive skills and reflecting on their own errors.

Several young children interacting with the robot

Understand the project in 4 minutes

How does it works?

A child working with the robot

As you can see in the video above, the children and the robot interact through a tactile tablet: the child prepares a word with small magnet letters, the robot write the word on the tablet, and the child correct the bad letters with the tablet’s stylus, either by rewriting the whole word, or only specific letters.

An example of interaction on a tactile tablet

When the child is satisfied, he moves on to the next word. The picture below shows an example of such a short text, written by the robot with the help of a 6 years old “teacher”:

 A example of short text written by the robot

Under the box, the robot runs an algorithm known as a Principal Component Analysis that allows to automatically identify the main differences within a dataset of letter’s samples. By manipulating the values of these differences, the robot can generate purposefully deformed letters, or letters that converge towards the ones demonstrated by the child, thus effectively “learning” from the child.

9 'g' generated by varying eigen values  obtained from a PCA decomposition

Academic Publications

A. Jacq; S. Lemaignan; F. Garcia; P. Dillenbourg; A. Paiva : Building Successful Long Child-Robot Interactions in a Learning Context. 2016. 11th ACM/IEEE International Conference on Human-Robot Interaction (HRI), Christchurch, NEW ZEALAND, MAR 07-10, 2016. p. 239-246.
S. Lemaignan; F. Garcia; A. D. Jacq; P. Dillenbourg : From Real-time Attention Assessment to “With-me-ness” in Human-Robot Interaction. 2016. 11th ACM/IEEE Conference on Human-Robot Interaction, Christchurch, New Zealand, 2016. p. 157-164.
A. D. Jacq; s. Lemaignan; F. Garcia; P. Dillenbourg; A. Paiva : Building Successful Long Child-Robot Interactions in a Learning Context. 2016. 11th ACM/IEEE Conference on Human-Robot Interaction, Christchurch, New Zealand, 2016. p. 239-246.
D. Hood; S. Lemaignan; P. Dillenbourg : The CoWriter Project: Teaching a Robot how to Write. 2015. 2015 Human-Robot Interaction Conference, Portand, USA, March 3-5, 2015. DOI : 10.1145/2701973.2702091.