A team of Japanese scientists at Kobe University, lead by Yukio-Pegio Gunji, have built a computer out of live crabs. Before you imagine a bunch of crabs shoved into a laptop casing, the reality is that the researchers were able to utilize the predictable nature of soldier crab swarms to build logic gates, the basic AND/OR problems that are the building blocks of computing. The team published a paper on their findings in the journal Complex Systems (PDF).
As reported by Gizmag, the behavior that the scientists were exploiting was what happens when two swarms of soldier crabs coming from different directions meet up: they meld into one another and operate as one, heading in the same direction that's at the combined angle of where they came from. The scientists used this behavior to design two different logic gates.The first, shown below, is a basic OR gate that consists of two legs that lead into one third leg. If either one swarm OR two are placed in the maze, in both scenarios the crabs head out the maze through the third leg. If either one swarm OR another swarm are placed in the maze, each will head out the third leg.
The second maze, shown below, is a basic AND gate that is shaped like an "X" with a fifth leg heading straight out of the top from the intersection of the other two legs. If only one swarm goes into the maze, it exits directly out the opposite end, but if one swarm AND another are placed the maze, when they meet, they'll head out the fifth leg in the center.
Here's what the actual soldier crab computer looks like.
It may not look like much, but using these basic principles and the predictable nature of the crabs' swarm behavior, it could theoretically result in more complex computers. Will that happen? It's unlikely. Crabs are not really what people are looking for in a computer. But this type of nature-as-computer modeling is really most useful to help inspire better technology, like how slime mold's ability to find the quickest path between two points is being used to come up with better transportation routes. Sometimes natural processes are just as smart as (if not smarter than) those created by humans.