kitchen table math, the sequel: travelling salesman

Wednesday, October 27, 2010

travelling salesman

solved by bees


Katharine Beals said...

If a colony of bees can do this, just imagine what a colony of humans could do.

Anonymous said...

I think I might wait until we find out how many destinations the bees needed to solve. If the number is small, then I'm not super impressed ... especially as the articles claim that the basic point was to see if the bees would just travel in the order the flowers were discovered.

If the bees can solve arbitrary 20 flower arrangements, then I'll be impressed. I suspect that they don't/can't ... partially because they don't *NEED* to.

Solving for 3-5 flowers (or flower patches) seems quite manageable/reasonable for bees. Solving for 20 does not.

Anyone have a link to the actual paper?

-Mark Roulo

Anonymous said...

I agree with Mark R. - unless there's an actual paper cited that actually says individual bees can find optimal routes among large sets of flowers, this is almost certainly another case of "science journalism" showing very little of either.

There's a cool clip from BBC's Animal Camera on bees where they put little radar reflectors on bees and then tracked them as they visited flowers. The tracks shown seem to be more like "fly out to a flower-rich area, wander around, fly back" with opportunistic feeding along the way (and especially during the "wander around" phase).

Anonymous said...

Maybe they are calculating distance by sense of smell (closer routes would have a stronger smell). Of course, this wouldn't help with solving a typical map problem.