Design Learning: a methodology for the autonomous design and manufacture of customised toys based on Machine Learning

Abstract

Abstract Society is increasingly demanding products with a higher degree of customization and shorter manufacturing phases. This demand is currently conditioned by a context of reduced mobility due to Covid-19. For these reasons, a new methodology called Design Learning is proposed, capable of designing personalized toys autonomously to improve the physical and mental development of children, including those with special needs. This method is based on the automated study of the user through their activity on audio-visual platforms and social networks, and on the application of Machine Learning techniques, Augmented Reality, Mixed Reality, parametric CAD and Additive Manufacturing. The results of the experiment show that children are able to use the proposed method and enjoy using it. It has also been obtained, thanks to the techniques used, in addition to designing and manufacturing toys autonomously and contributing to the development of children, that this tool also favours the competence of small companies and reduces the carbon footprint produced by traditional manufacturing processes.