Williams was one of three operators of the parts retrieval system, a five-story robot built by the Unit Handling Systems division of Litton Industries. The robot was designed to retrieve castings from high density storage shelves at the Flat Rock plant. Part of the machine included one-ton transfer vehicles, which were carts on rubber wheels equipped with mechanical arms to move castings to and from the shelves. When the robot gave erroneous inventory readings, Williams was asked to climb into the racks to retrieve parts manually. Another news account states the robot was not retrieving parts quickly enough.

He climbed into the third level of the storage rack, where he was struck from behind and crushed by one of the one-ton transfer vehicles, killing him instantly. His body remained in the shelf for 30 minutes until it was discovered by workers who were concerned about his disappearance.

His family sued the manufacturers of the robot, Litton Industries, alleging “that Litton was negligent in designing, manufacturing and supplying the storage system and in failing to warn [system operators] of foreseeable dangers in working within the storage area.” In a 1983 jury decision, the court awarded his estate $10 million and concluded that there simply were not enough safety measures in place to prevent such an accident from happening. He would go down in history as the first recorded human death by robot. The award was raised to $15 million in January 1984. Litton settled with the estate of Williams for an undisclosed amount in exchange for Litton not admitting negligence.

  • FooBarrington@lemmy.world
    link
    fedilink
    English
    arrow-up
    4
    ·
    1 year ago

    Part of it is probably wrong extrapolation from experience. Even if I try to punch someone as hard as I can, there’s a comparatively low limit of damage I can do. Sure, in the wrong spot I could kill someone or seriously hurt them - but my body severely limits what I can do so I don’t hurt myself.

    Robots don’t have these natural limits. They can be programmed in, but code can always fail. A robot doesn’t care if it will hurt itself by accelerating too quick or by continuously exerting high amounts of force. It’s as mindless as a rock crushing somebody, and much more agile.