The major inspiration for the self-repairing concept is biological cell development: the formation of patterns from undifferentiated masses. A number of computational systems in areas such as computer graphics and learning neural networks have been created in association with the ideas of biological development. Self-repair systems are of significant importance for the design of computational systems. Yet it is less certain whether local chemical interactions can actually give rise to various patterns in biology.

The paper considers self-repair as an intrinsic part of regular functioning, avoiding the difficulties of explicit fault detection. This process is implemented by a cellular automaton construction where pattern reconfigurations depend on changes of neighboring states. The update rules use the Cartesian genetic programming method with a fitness function providing fault-tolerant self-assembly. It is realized that joint bulk computations are not feasible without establishing a priority order. Thus, this work utilizes a raster scan that, starting at the upper-left corner, produces some order for possible optimization.

Ordinarily neglected, such a prioritization was also a crucial factor for the operability of my model [1].