Fwd: Re: robot fly

From: Dale K Stalnaker (Dale.K.Stalnaker@grc.nasa.gov)
Date: Mon Aug 19 2002 - 15:48:15 EDT

  • Next message: bivalve: "Re: Orr on No Free Lunch"

    Here's more detail about the "Robot Fly".

    Dale Stalnaker

    >X-Sender: klimek@parrot.grc.nasa.gov
    >X-Mailer: QUALCOMM Windows Eudora Pro Version 4.2.2
    >Date: Mon, 19 Aug 2002 14:54:30 -0400
    >To: Dale K Stalnaker <Dale.K.Stalnaker@grc.nasa.gov>
    >From: Bob Klimek <klimek@grc.nasa.gov>
    >Subject: Re: robot fly
    >
    >Dale,
    >
    >I found the article in New Scientist. It gives a little more detail.
    >
    >http://www.newscientist.com/news/news.jsp?id=ns99992673
    >
    >Bob
    >
    >
    >At 12:06 PM 8/19/2002 -0400, you wrote:
    >>Bob
    >>
    >>Thanks, this is a fascinating article. Have you found any other
    >>references to the same experiment?
    >>
    >>
    >>At 10:11 AM 8/19/2002 -0400, you wrote:
    >>>Dale,
    >>>
    >>>Here's one of the articles I found.
    >>>
    >>>http://www.reuters.com/news_article.jhtml?type=search&StoryID=1329646
    >>>
    >>>Bob
    >>
    >>
    >>There are a few things I don't understand. The robot had one goal--to
    >>produce maximum lift. It had no pre-programmed concept of flapping and
    >>eventually learned that maximum lift could be attained by flapping its
    >>wings. The last paragraph tells us that it was too heavy to actually fly.
    >>
    >>Now since it could not _really_ fly, then what would motivate it to try
    >>to "cheat" by standing on its wingtips and climbing on objects? The
    >>article doesn't tell us exactly _how_ it knows it has achieved it's
    >>goal. If the goal was to achieve maximum altitude, then cheating (ie,
    >>climbing) would be the most successful strategy, rather than
    >>flapping. So WHY did it finally decide that flapping was _better_than
    >>cheating, since flight wasn't really possible?
    >>
    >>Perhaps someone intervened and programmed it to use another method.
    >>
    >>Now on the other hand, let's assume it could actually fly. And let's
    >>also give it another goal--to escape from a predator. I would think it
    >>would select flapping as the best method. From a darwinian perspective
    >>this would be the best survival strategy.
    >>
    >>Dale
    >>
    >> >>>
    >>>Cheating was one strategy tried and rejected during the process of
    >>>artificial evolution -- at one point the robot simply stood on its wing
    >>>tips and later it climbed up on some objects that had been accidentally
    >>>left nearby.
    >>>
    >>>But after three hours the robot discovered a flapping technique --
    >>>rotating its wings through 90 degrees, raising them, then twisting back
    >>>to the horizontal before pushing back down.
    >>>
    >>>"This tells us that this kind of evolution is capable of coming up with
    >>>flying motion," said Peter Bentley, an evolutionary computer expert at
    >>>University College, London.
    >>>
    >>>However, the robot could not actually fly because it was too heavy for
    >>>its electrical motor.
    >> >>>
    >
    >

    ----------------------------------------------
    Dale K. Stalnaker
    NASA/Glenn Research Center
    Power & Propulsion Office
    dale.k.stalnaker@grc.nasa.gov
    PHONE: (216) 433-5399
    FAX: (216) 433-2995



    This archive was generated by hypermail 2.1.4 : Mon Aug 19 2002 - 18:18:17 EDT