From human creative cognitive processes to adaptable artificial system design

In epigenetic robotics, a new research field, interdisciplinary theory and empirical evidences are used to inform adaptive robotic models, and, vice-versa, these models can be used as tools to make experimental predictions in developmental psychology. A truly autonomous robot should be capable of evolving and `growing up' through experience. Following different psychological points of view, growing up implies a knowledge creative process called: change of functional meaning; increased complexity; enlargement of the internal knowledge map; abstraction and insight. To understand this creative process, we organized an experiment with pre-school children diving with the abstraction process. The cognitive development of children of this age do not include the ability of abstraction, but they are able to explain the process they are thinking. Forty-two metaphoric sentences have been proposed to eight working-groups, of nine to ten children each, asking for their abstracted meanings. After a preliminary "brainstorming" phase , where the free creative associations were prevalent, we addressed children's attention towards the individuation of the metaphoric meaning. The process has been recorded and then we analyzed and classified the answers. Collective speech have been analyzed to compensate the individual differences. The children disclosure was mostly driven by their value system, their motivations and their emotions. They tried many different strategies to reach the abstract meaning, starting from their concrete knowledge and experiences. Each children followed a set of thinking paths that resulted in some very interesting suggestions for the architecture of an adaptive and evolving robot: i.e. the importance of multi-sensor perception, motivation and emotional drives are underlined and, the growing up insights shows similarities to emergent self-organized behaviours.