Researchers are taking inspiration from nature to design the most efficient and repairable power and information networks, where essential redundancy patterns resemble the complex and beautiful branching of snowflakes.
Redundancy is used in networks for the sake of being able to repair on the fly. traditionally, if a power grid is damaged, it will entirely shut down, or will at least have to be shut down in order to be repaired. The same goes for information pathways, where everything could stop if a sole path is interrupted.
However, if there are multiple and redundant pathways, damage may not necessarily interrupt the whole system, allowing for repair even as the network still functions.
This concept can be seen in roads, today. Even if an accident or natural disaster cuts off one road, there are usually a number of detours a traveler can take to still get where they need to go. Redundancy eliminates the danger of choke points.
However, in that same breath, engineers and manufactures can argue that redundancy can be costly and inefficient.
That's why Robert Farr of the London Institute for Mathematical Sciences recently investigated which network structures were most efficient in maintaining benefits of redundancy while simultaneously preserving as much resource efficiency as possible. He quickly found that nature, as should be expected, is still the best engineer. (Scroll to read on...)
According to a study recently published in the Physical Review Letters of the American Physical Society (APS), the best networks are made from partial loops around the units of a grid, with exactly one side of each loop missing. All of these partial loops link together, back to a central source, making the whole system exceptionally resistant to multiple breaks. Amazingly, if you tried to plot these kind of networks out, they look remarkably like snowflakes - which naturally branch in symmetrical and highly efficient ice crystals.
Farr and his colleagues are now looking into how these 'snowflake' network designs can be implemented to boost the efficiency of real-world power grids and information pathways.