A New Paradigm for Developing Digital Systems Based on a Multi-Cellular Organization

Embryological electronics or “Embryonics” is a new paradigm for developing digital systems of any complexity, endowed of universal computation, self-repair and self-reproduction capabilities. Embryonics, which can also be defined as a quasi-biological development of atificial systems, is based on the natural mechanism of development of living multi-cellular beings. Starting with a single mother cell containing the description of the organism in the form of a genome, the final organism is achieved through a succession of cell divisions, occurring with the differentiation of each cell, i.e. a specialization dependent essentially on the physical position of the cell (i.e. on its coordinates) in the given space. Applying this technique to the design of digital neural networks could lead to the development of artificial organisms simulating both the biological dcvelopment and the final architecture of natural neural networks.

• Type: Conference
• Year: 1995
• Where: ISCAS
• Authors: Daniel Mange, Serge Durand, Eduardo Sanchez, André Stauffer, Gianluca Tempesti, Pierre Marchal, Christian Piguet
• Keywords: Digital Neural Networks, Hardware, ISCAS 1995, Neural Networks, Single Mother Cell