Slashdot Mirror


Google Patents Self-Driving Car That Glues Pedestrians To The Hood In A Crash (cnn.com)

An anonymous reader quotes a report from CNN: Google just got a patent for a special kind of coating on self-driving cars that could help prevent pedestrian injuries. The company wants to coat autonomous vehicles with a sticky substance so that if they hit a pedestrian, the person would be glued to the car instead of flying off. "[The pedestrian] is not thrown from the vehicle, thereby preventing a secondary impact between the pedestrian and the road surface or other object," says the patent, granted on Tuesday. Google explains that an "adhesive layer" would be placed on the hood, front bumper and front side panels of a car. A thin coating would protect it until an impact occurred. Google is paying Arizona residents $20 per hour to test its self-driving vehicles.

1 of 203 comments (clear)

  1. Re:physics! by rgbatduke · · Score: 3, Interesting

    Super-good physics. Now consider the practical aspect of driving a car with active glue all over the paint through:

    * Bugs. I can coat my windshield with bug guts in a single night drive from my house to the coast through various swampy regions. And my windshield isn't sticky.

    * Pollen. I live in NC, and in the spring my car -- well, really everything, not just my car, inside and out -- is coated with a layer of tree pollen so thick that it is virtually opaque. Again, this sticks to hard finish PAINT. I can only imagine how tightly it will adhere to glue as it sits out in the sun.

    * Dust. Even when the pollen season is over, there is always dust in the air. Always. That's why we have to wash our cars a few times a year unless we want people to write "wash me" with their fingers on our fenders and windshields.

    * Salt + grime. Yes I live in the south without that problem so much but h/t to our northern cousins who have to drive through slushmelt. I'm guessing immersing an active glue surface in slushmelt would pretty much end the effectiveness of the glue.

    * Fall leaves. For that matter, summer leaves if you park under live oaks or evergreens. Maple seeds. Acorns. Birds. Squirrels. I can't wait to come out some morning and see my self-driving car with a seagull glued to the front fender. Oh my.

    * Children, pets, old people.

    I would say that all of this makes it impractical to drive a car with active sticky flypaper instead of paint as a "permanent" primary front coating. Laughably impractical.

    Which leaves us with the only alternative -- a "glue bag" as a sort of external equivalent of an internal airbag. Now, is it possible to come up with a formula for stickum that can be sprayed in the (say) millisecond before a collision so that it coats a pedestrian -- but not their mouth or eyes or lungs, which would be "bad" -- and the front bumper just in time to catch the human and stick them and hold during the millisecond or so that they are actually in contact with the bumper/hood? Personally I doubt it. I know of no glues that can be applied and will stick and set in a millisecond, especially glues that are non-toxic and safe to spray onto random humans to lower risk of death. Can you imagine an explosion of super-glue all over somebody -- not that super-glue can come close to bonding in a millisecond.

    So this seems like a really stupid idea too. Which makes the entire idea sound incredibly stupid, not worth the money required to patent it. If they wanted to accomplish the same thing in a PRACTICAL way, they could just mount airbags on the front fender that were triggered by certain conditions, such as an impending collision. That would actually be USEFUL -- and not just for humans. Having a heavy-duty airbag go off to cushion a regular collision between two cars could actually significantly reduce the average force during the impulse by spreading it out over a meter BEFORE starting to crumple the front bumper accordion. It would also do exactly the same thing as the glue to a pedestrian only better -- catch them on a meter or so of compressing air while the impulse matches their speed to that of the car. You might even be able to make the bag itself "sticky", although I suspect that would interfere with its explosion -- at least you might be able to make it out of e.g. neoprene with a non-stick stickiness.

    Not impressed, wouldn't invest.

    --
    Even when the experts all agree, they may well be mistaken. --- Bertrand Russell.