Key words :
  Self-replication, self-repair, John von Neumann, cellular automata, embryonics,
  universal computation, universal construction, Turing machine.

  Project leader :
  D. Mange

  Participation :
  D. Mange, A. Stauffer, G. Tempesti, F. Restrepo, E. Petraglio, L. Prodan

  Description (goals, methods, perspectives) :
  In the 1950s, John von Neumann designed a cellular automaton capable of
  self-replication and satisfying two strict conditions:
  (1) universal computation, i.e., the ability to realize a universal Turing machine;
  (2) universal construction, i.e., the ability to construct a digital system of any given
  size. In spite of the efforts of several researchers over the years and the considerable
  technological advances, even today the complete hardware implementation of von
  Neumann's original automaton remains beyond our reach, and only partial simulations
  have been realized. This failure is due to the extreme complexity of the project and to
  the difficulty of satisfying the two conditions of construction and computation
  universality.

  The future objectives of our project are:
  - the complete hardware implementation of a self-replicating automaton, which
  entirely fulfills von Neumann's two original conditions: universal computation and
  universal construction;
  - the introduction of two novel features to the self-replicating automaton: self-repair
  and the capability of real-time operation (i.e., a reactive system).

  Self-repair is indispensable to assure the continuous and reliable functioning of a
  complex hardware system. The ability to function in real time will allow the final
  automaton to be used in an industrial computing system, which interacts with its
  environment.

  The fundamental motivation of this project is first and foremost scientific in nature:
  finding an answer to a major, hitherto unsolved problem in computer science. There
  are also technical motivations, stemming from the desire to develop self-repairing
  computer hardware (very large scale integrated circuits) that will be at the heart of
  tomorrow's computing industry.

  Main results during this year :
  The development of a new molecule (MuxTree), the basis of a programmable logic
  circuit, has allowed us to introduce the automatic detection of faults, indispensable for
  self-repair.
  A first prototype of a universal Turing machine is under development.
 
 

 Main publications :

  B. Girau, P. Marchal, P. Nussbaum, A. Tisserand, H. F. Restrepo

  Evolvable Platform for Array Processing: A One-Chip Approach

  Proceedings of the Seventh International Conference on Microelectronics for Neural, Fuzzy and
  Bio-Inspired Systems, MicroNeuro '99, Granada, April 7-9 // pp. 187-193 ; (1999).

  G. Tempesti, D. Mange, A. Stauffer

  Embryonics: Multi-Cellular and Multi-Molecular Digital Systems

  Half-day Colloquium Evolutionary Hardware Systems, IEE, London, 2 June // pp. 1/1-1/4 ;
  (1999).

  G. Tempesti, D. Mange, A. Stauffer

  The Embryonics Project: A Machine Made of Artificial Cells

  Rivista di Biologia, Biology Forum // Vol. 92, No 1, pp. 143-188 ; January-April (1999).

  A. Stauffer, M. Sipper

  On the relationship between cellular automata and L-systems: The self-replication case

  Physica D // 116, pp. 71-80 ; (1998).

  A. Stauffer, M. Sipper

  Modeling Cellular Development Using L-Systems

  In M. Sipper, D. Mange, A. Pérez-Uribe, editors, Evolvable Systems: From Biology to Hardware,
  volume 1478 of Lecture Notes in Computer Science, Springer, Berlin // pp. 196-205 ; (1998).

  D. Mange, A. Stauffer, G. Tempesti
  Embryonics: A Microscopic View of the Molecular Architecture

  In M. Sipper, D. Mange, A. Pérez-Uribe, editors, Evolvable Systems: From Biology to Hardware,
  volume 1478 of Lecture Notes in Computer Science, Springer, Berlin // pp. 185-195 ; (1998).

  D. Mange, A. Stauffer, G. Tempesti

  Embryonics: A Macroscopic View of the Cellular Architecture

  In M. Sipper, D. Mange, A. Pérez-Uribe, editors, Evolvable Systems: From Biology to Hardware,
  volume 1478 of Lecture Notes in Computer Science, Springer, Berlin // pp. 174-184 ; (1998).

  D. Mange, E. Sanchez, A. Stauffer, G. Tempesti, P. Marchal, C. Piguet

  Embryonics: A New Methodology for Designing Field-Programmable Gate Arrays with
  Self-Repair and Self-Replicating Properties

  IEEE Transactions on VLSI Systems // Vol. 6, No 3, pp. 387-399 ; September (1998).

  D. Mange, M. Sipper

  Von Neumann's Quintessential Message: Genotype + Ribotype = Phenotype

  Artificial Life // Vol. 4, No 3, pp. 225-227 ; Summer (1998).

  D. Mange, M. Tomassini (eds)

  Bio-Inspired Computing Machines

  Presses polytechniques et universitaires romandes, Lausanne ; (1998).

  G. Tempesti

  A Self-Repairing Multiplexer-Based FPGA Inspired by Biological Processes

  EPFL, Lausanne // Thèse No 1827 ; (1998).

  G. Tempesti, D. Mange, A. Stauffer

  Il progetto Embryonics: una macchina fatta di cellule artificiali

  Systema Naturae, Annali di biologia teorica // No 1, pp. 41-82 ; (1998).

  G. Tempesti, D. Mange, A. Stauffer

  Self-Replicating and Self-Repairing Multicellular Automata

  Artificial Life // Vol. 4, No 3, pp. 259-282 ; Summer (1998).

  M. Sipper

  Fifty Years of Research on Self-replication: an Overview

  Artificial Life // Vol. 4, No 3, pp. 237-257 ; Summer (1998).

  M. Sipper, D. Mange, A. Pérez-Uribe (eds)

  Evolvable Systems: From Biology to Hardware ,

  Lecture Notes in Computer Science 1478, Springer, Berlin ; (1998).

  M. Sipper, D. Mange, A. Stauffer

  Ontogenetic Hardware

  BioSystems // Vol. 44, No 3, pp. 193-207 ; (1997).

  M. Sipper, E. Ruppin

  Co-evolving architectures for cellular machines

  Physica D // 99, pp. 428-441 ; (1997).