The Pigeon Guidance System

Discussion in 'The Bench' started by Buick_350X, Mar 9, 2003.

  1. Buick_350X

    Buick_350X Guest

    The Pigeon Guidance System


    During WWII, the US military funded a program that researched a missile
    guidance system that used a pigeon to direct a rocket towards enemy boats
    and submarines. They were trained inside of a "Skinner box" and given food
    rewards (reinforcement) for correctly steering the missile in the right
    direction. If there were no ships in sight, they would simply keep the
    missile flying straight out over the ocean. When they did encounter an
    object, they would then peck at its location and guide the missile directly
    at it. This device actually worked and with amazing accuracy.

    *****

    In one of the odder stories of World War II, the US military backed an
    investigation into weapons guided by trained pigeons. The investigation was
    conducted by the well-known behavioral psychologist Burrhus Frederic
    Skinner.
    In early 1942, Skinner, then of the University of Minnesota, conducted
    preliminary studies on the concept of using trained animals as a guidance
    system.

    The US National Defense Research Committee (NDRC), which funded new
    technologies that might be useful for winning the war, was skeptical of the
    idea, but in mid-1943 awarded a $25,000 USD contract to General Mills to
    continue the work. The investigation was codenamed PROJECT PIGEON, and
    Skinner hoped to be able to use pigeons to guide a weapon to within 6 meters
    (20 feet) of a target.

    Three pigeons were each tucked into a jacket made of a sock and then put
    into a harness inside the guidance system, facing a screen. An image of the
    target was projected onto each of the three screens through a lens system in
    the nose of the weapon, with crosshairs defined by beams of light. Each
    pigeon was supposed to peck at its screen, which was wired to provide
    feedback to the missile's flight controls, to keep the crosshairs on target.
    The system accepted inputs from all three pigeons, but only acted if two or
    all three agreed.

    The pigeons were trained with slides of aerial photographs of the target,
    and if they kept the crosshairs on the target, they were rewarded by a grain
    deposited in a tray in front of them. Skinner later found that the pigeons
    were less easily disturbed under confusing circumstances if they were fed
    hemp (marijuana) seeds rather than grains.

    Skinner hoped to fit the pigeon guidance system to a Pelican, but he never
    managed to overcome official skepticism. When he put on a demonstration in
    New Jersey of the pigeon guidance system for officials of the US Office of
    Scientific Research and Development (OSRD) he was bitterly frustrated to see
    they were amused rather than impressed.

    PROJECT PIGEON was abandoned. Skinner went home with 24 trained pigeons,
    which he kept in a dovecote in his garden. Whether the idea was practical or
    not, it appears that Skinner as a academic psychologist simply was not on
    the same wavelength as the industrial engineers and military officials he
    was trying to work with, and never managed to communicate with them
    effectively.
     

Share This Page