Forgot your password?
typodupeerror
Science Technology

Slime Mold Could Lead To Better Tech 179

Posted by ScuttleMonkey
from the expensive-mold-or-cheap-trains dept.
FiReaNGeL writes to tell us that recent observation of slime mold could eventually lead the way to improved tech like better computer and communications networks. "This revelation comes after a team of Japanese and British researchers observed that the slime mold connected itself to scattered food sources in a design that was nearly identical to Tokyo's rail system. Atsushi Tero from Hokkaido University in Japan, along with colleagues elsewhere in Japan and the United Kingdom, placed oat flakes on a wet surface in locations that corresponded to the cities surrounding Tokyo, and allowed the Physarum polycephalum mold to grow outwards from the center. They watched the slime mold self-organize, spread out, and form a network that was comparable in efficiency, reliability, and cost to the real-world infrastructure of Tokyo's train network."
This discussion has been archived. No new comments can be posted.

Slime Mold Could Lead To Better Tech

Comments Filter:
  • Fred Physarum (Score:5, Insightful)

    by Drantin (569921) on Friday January 22, 2010 @03:17PM (#30862348)

    To think. After all these years, Fred Physarum [gpf-comics.com] is finally getting the recognition he deserves.

  • by wizardforce (1005805) on Friday January 22, 2010 @03:19PM (#30862372) Journal

    Maybe the slime mold has been evolving for millions of years and there really isn't much in the way of improvements that can be made.

  • by Jeng (926980) on Friday January 22, 2010 @03:21PM (#30862398)

    It's a smart scientist who does not re-invent the wheel.

  • by natehoy (1608657) on Friday January 22, 2010 @03:48PM (#30862664) Journal

    Using slime mold to study two other kinds of slime seems either redundant or self-evident. I can't decide which.

  • Re:uh.. (Score:3, Insightful)

    by interkin3tic (1469267) on Friday January 22, 2010 @04:02PM (#30862844)

    No, the people who designed the tokyo metro layout [mkc-properties.com] on the other hand, were most definitely high on something.

    "Let's design an extremely interconnected subway system, except that in order to get across downtown you need to change lines 3 times including once to a completely different rail system. Some areas of downtown will have a train station every half block, wheras others will be mostly empty. To balance out that inconvinience, lets make all the trains run on time down to about 3 seconds, have all the stops in at least two languages, and keep the stations cleaner than most resturaunts."

  • by John Hasler (414242) on Friday January 22, 2010 @04:42PM (#30863176) Homepage

    He can certainly equal the performance of the average day-trader.

  • I hate slime molds (Score:3, Insightful)

    by tsotha (720379) on Friday January 22, 2010 @05:19PM (#30863586)
    They do 1D6 of Constitution damage and there's no way to get it back.
  • by julesh (229690) on Friday January 22, 2010 @05:31PM (#30863762)

    the proper conclusion is that japanese transportation engineers are no smarter than slime molds

    Or indeed soap, which is also able to perform similar optimization tasks: take two pieces of perspex and join them together using bolts arranged in the pattern of your major destinations with a gap of around 1-2cm between them. Dip in a strong soap/water mixture and remove carefully. You should find a series of large bubbles have formed, with edges running between the bolts. Surface tension will probably have resulted in those edges being an optimal or close to optimal solution to the problem of joining them together with the most efficient network. Repeat several times, take the most common result.

    Simple energy reduction problems like this aren't a useful test of anything. There are plenty of natural processes that don't involve intelligence that are more than capable of solving them.

  • by jsveiga (465473) on Friday January 22, 2010 @05:55PM (#30864080)

    I assume the mold paths solution simply "converged" to the most efficient way of carrying the nutrients between the nodes. As it was mentioned here, soap bubbles will also "find" the shortest paths, as will the mold's "brute force" approach (broad spread, then coalesce to the most efficient ones).

    But the natural solutions would not take into account the human distribution and convenience, as each node (apart from the big central oat flake) have the same appeal to the mold - and possibly the ones closest to the borders have less appeal (or more "cost"). Same goes for the surface tension solution (soap).

    What if the human factor shifts the "weight" of some nodes and paths? For example, there might be very few people needing to go from node A to B, but many needing to go from A to C, so although a "natural" solution would only take the distances and positions into account, a "human" solution would want to favor the trip from A to C even if that meant making the A-B trip worse.

    So if the mold solution is really very similar to the real rail system, then either Japanese commuters are amazingly "natural" in regards to where they live, where they work, and demographic distribution, or the Japanese railroad engineers missed the human factor when designing the grid. The first possibility is somehow beautiful and creepy at the same time.

  • by Seor Jojoba (519752) on Friday January 22, 2010 @06:17PM (#30864386) Homepage

    It's a little sad that somebody, in pursuit of an audience, had to angle the story towards "we could be using mold to make design decisions." Your mass transit planners are not going to call in a consultant with a suitcase full of mold, obviously. The paths chosen for rail have so many political factors that the "most efficient" model has little relevance.

    But just stop thinking of utility for a moment. Look at those pictures of the mold growing to reach all points and form little roads between them. That is fantastic! "Because you could then plan light rail and freight logistics and--" STOP! No, don't jump on to the practical applications yet. Take a moment to think about that simple little organism doing that complicated thing and how cool that is. Those pictures are breathtaking.

    And after that, maybe try to write a matching algorithm to see if you can predict which paths will form by the slime. And then see if that algorithm offers something that the human-designed ones don't have already. And then maybe integrate and devise new algorithms based on what was learned. And then see what practical applications there are for these algorithms. This is what the scientists and engineers will actually end up doing if it is possible. Can we stop acting like bored little brats that every scientific observation isn't immediately useful?

  • by sznupi (719324) on Friday January 22, 2010 @06:26PM (#30864490) Homepage

    Perhaps not quite. There are beetles forming spherical "boulders" of organic matter, that's quite close to wheel conceptually. Spherical plants moved by wind. And you can find even closer analogues in microorganisms...

    The main problem with evolving large scale "proper" wheel, I guess, is of intermediate structures; apparently they were worse for survival then the alternatives.

Machines certainly can solve problems, store information, correlate, and play games -- but not with pleasure. -- Leo Rosten

Working...