Denne artikel fra Science synes jeg er ret fascinerende; forskere har lagt havreflager ud i et mønster som svarer til placeringen af "points of interest" i Tokyo og omegn, og slimsvampen
P. polycephalum forbinder dem i løbet af et døgns tid med et meget effektivt netværk, som i mange tilfælde ligner subway-systemet i Tokyo og omegn - og i de tilfælde hvor det ikke gør, er det lige så effektivt. Det er problemer som man næppe kan finde nogen god matematisk løsning på, så jeg synes det er ret interessant at noget som en slimsvamp kan finde frem til tæt på optimale løsninger.
Ride the Slime Mold Express!Science skrev:
By Tim Wogan
ScienceNOW Daily News
21 January 2010
If you want to design a railway system, you could do worse than hire a slime mold. Researchers have shown that, when grown on a map of Japan, the gelatinous, funguslike organism connects points of interest in a pattern similar to Tokyo's train network. Engineers might be able to take a cue from the organism's approach to design more-efficient transportation systems.
The trick has to do with how slime molds eat. When Physarum polycephalum, a slime mold often found inside decaying logs, discovers bacteria or spores, it grows over them and begins to digest them through its body. To continue growing and exploring, the slime mold transforms its Byzantine pattern of thin tendrils into a simpler, more-efficient network of tubes: Those carrying a high volume of nutrients gradually expand, while those that are little used slowly contract and eventually disappear.
Researchers have harnessed this behavior to amusing effect in the past. In 2000, for example, a team led by mathematical biologist Toshiyuki Nakagaki of Hokkaido University in Japan, showed that P. polycephalum could find the shortest path through a maze to connect two food resources. (The work won an Ig Nobel prize.)
But that was a puzzle with a single correct solution. In the new work, the team wanted to know how the mold would perform in a real-world situation in which several competing objectives had to be balanced at once. Designing a railway network that connects many cities presents just such a problem. "The planning is very difficult because of the tradeoffs," says cell biologist Mark Fricker of the University of Oxford in the United Kingdom, who was also involved in the research. For example, connecting all cities by the shortest possible length of track often compels travelers to take highly indirect routes between any two points and can mean that a single failure isolates a large part of the network. Building in more redundancy makes the network more convenient and more resilient, but at a higher cost.
Because they couldn't mathematically determine a "perfect" solution, the researchers decided to task the slime mold with a problem human designers had already tackled. They placed oat flakes (a slime mold favorite) on agar plates in a pattern that mimicked the locations of cities around Tokyo and impregnated the plates with P. polycephalum at the point representing Tokyo itself. They then watched the slime mold grow for 26 hours, creating tendrils that interconnected the food supplies.
Different plates exhibited a range of solutions, but the visual similarity to the Tokyo rail system was striking in many of them, the researchers report in tomorrow's issue of Science. Where the slime mold had chosen a different solution, its alternative was just as efficient.
If researchers could construct a computer model of the slime mold's behavior, says Fricker, it might help engineers design better transportation networks. "The idea would be that, if one put it into a new context, a system using these rules would build a network that ought to have respectable properties."
The work is "a very interesting example of how biology can inspire new methods in technological design," says Melanie Mitchell, a computer scientist at Portland State University in Oregon. But she's not quite ready to jump on the slime mold express. "This paper uses only one relatively simple example," she cautions. "It's not obvious that similar experiments would work as well for matching other transport networks."
Jeg har ikke læst hele artiklen endnu, kun overblik som ovenstående, men synes konceptet er ret fantastisk.