In 1948, von Neumann set out to describe a model of a self-reproducing machine in a paper called “The General and Logical Theory of Automata” that he wrote for the Hixon Symposium. He had not yet conceived of cellular automata and could not completely solve the problem of how, in theory, a machine could self-reproduce. Only after the suggestion by his colleague Stanislaw Ulam to use a cell-based concept, was von Neumann able to formulate a model for a machine that was fully capable of self-reproduction.
The theory of automata is a relatively recent and by no means sharply defined area of research. It is an interdisciplinary science bordered mathematically by symbolic logic and Turing machine theory, bordered engineering-wise by the theory and the use, par- ticularly for general non-numerical work, of large scale computing machines, and bordered biologically by neurophysiology, the theory of nerve-nets and the like. Problems range from Godel-type questions (relating to Turing machines and decision procedures), to questions of duplication, of various biological phenomena in a machine (e.g., adaptation, self-reproduction and self-repair)
One important part of von Neumann’s work on automata relates to the problem of designing reliable machines using unreliable components. Given a set of building blocks with some positive probability of malfunctioning, can one by suitable design construct arbitrarily large and complex automata for which the overall probability of incorrect output is kept under control?