5 ms·

Has to be very hard to separate signal from noise on the deployment of the air bag. Those dummies were in very controlled environments and the helmet deployed a

by hop 16y ago

Has to be very hard to separate signal from noise on the deployment of the air bag. Those dummies were in very controlled environments and the helmet deployed at very low rates of acceleration change. In the hit from behind demonstration, you can see the rider's head experienced only around 1g of acceleration and little jerk. It would get annoying as hell when this thing senses similar accelerations from non-car-hitting-you movements like bending over to roll up a pant leg or changing positions. Prankster friends would have a heyday too. Dainese has an has an impressive motorcycle airbag suit - http://www.youtube.com/watch?v=mo9Vlt5tGwY http://www.youtube.com/watch?v=mo9Vlt5tGwY tl;dr - That thing would go off all that time if they made it sensitive enough to work.

varjag 16y ago
The same problem was with early car airbags. The issue is long resolved.
- gjm11 16y ago
  It may be the same problem, but I wouldn't count on the same resolution. A car airbag deploys after the car has crashed into something but before that has a chance to injure the passenger, whereas in a bicycle accident the first impact could be with the cyclist's head. (Perhaps you can make airbag-helmets that inflate when the bike hits something, which will do for the case where you hit something in the road and get thrown off, or where a car collides with the bike. Not so good for tree branches and such, though.)
  - varjag 16y ago
    Yes, I'm pretty certain the rig uses some kind of accelerometers to detect the threshold event. Any fall from the seat is typically preceded by a kick. And yes, you can certainly can still "hit the ceiling", but it's not clear if such mode of injury is common among cyclists. By any measure, it sure beats riding without helmet.
    - gjm11 16y ago
      > it sure beats riding without helmet. Obviously an airbag helmet will be safer (but more inconvenient) than no helmet. Presumably it will be less safe (but maybe more convenient and better looking) than an ordinary helmet. It needs to beat both to succeed: that is, either some people who currently use helmets need to be convinced that the loss in safety from using an airbag helmet is small enough to be outweighed by the gain in convenience or looks or whatever, or some people who currently don't use helmets need to be convinced that the gain in safety from using an airbag helmet is large enough not to be outweighed by the loss in convenience or looks or whatever. This seems pretty difficult to me, especially as I bet the convenience gain isn't very big: I bet an effective airbag helmet needs to fit you about as well as an ordinary helmet, and it'll need to be put on with about as much care.
- hop 16y ago
  The acceleration your car experiences in an accident is an order of magnitude higher than anything your car ever experiences in everyday use, making the control system very easy to engineer - something like anything over 20g's and it pops (the algorithms take vehicle speed into account too for a varying threshold and probably use jerk.) In that video where the rider was hit from behind, there was very low relative change in velocity of the sensors in that helmet - no direct impact spike - and its on par with the same levels of acceleration you would experience in everyday bike riding.
  - varjag 16y ago
    No, it was not easy at all with the cars. You can easily have 20g impact when you speed over road bump. Also, I'm sure in the 5 years they spent on product R&D, the concerns that came up to you in few minutes here were raised too.
    - hop 16y ago
      You can easily have 20g impact when you speed over road bump. The accelerometers in your car are bolted to your 4,000lb frame, riding on dampened shocks and rubber tires - you will never experience acceleration anywhere close to what you would in a frame-on collision. They don't have a shipping product yet.
      - varjag 16y ago
        You will easily experience that, just that energies involved are minuscule compared to collision. The mass of your car frame is irrelevant, shockwaves still propagate through solid bodies. When someone slaps your face hard, that can easily be 100s of Gs without much ill effect.
forgottenpaswrd 16y ago
That's what I was thinking, e.g look at the impacts on mtb freestyle: http://www.youtube.com/watch?v=DDaqcEkQRnM&feature=related http://www.youtube.com/watch?v=DDaqcEkQRnM&feature=relat... You can use tracker http://www.cabrillo.edu/~dbrown/tracker/ http://www.cabrillo.edu/~dbrown/tracker/ to calculate the decelerations on jumps. They are huge, bigger than the car hitting the back on the bike. The only way to do this is to monitor the direction of the deceleration, not only the magnitude.

by hop 16y ago

varjag 16y ago

The same problem was with early car airbags. The issue is long resolved.

gjm11 16y ago
It may be the same problem, but I wouldn't count on the same resolution. A car airbag deploys after the car has crashed into something but before that has a chance to injure the passenger, whereas in a bicycle accident the first impact could be with the cyclist's head. (Perhaps you can make airbag-helmets that inflate when the bike hits something, which will do for the case where you hit something in the road and get thrown off, or where a car collides with the bike. Not so good for tree branches and such, though.)
- varjag 16y ago
  Yes, I'm pretty certain the rig uses some kind of accelerometers to detect the threshold event. Any fall from the seat is typically preceded by a kick. And yes, you can certainly can still "hit the ceiling", but it's not clear if such mode of injury is common among cyclists. By any measure, it sure beats riding without helmet.
  - gjm11 16y ago
    > it sure beats riding without helmet. Obviously an airbag helmet will be safer (but more inconvenient) than no helmet. Presumably it will be less safe (but maybe more convenient and better looking) than an ordinary helmet. It needs to beat both to succeed: that is, either some people who currently use helmets need to be convinced that the loss in safety from using an airbag helmet is small enough to be outweighed by the gain in convenience or looks or whatever, or some people who currently don't use helmets need to be convinced that the gain in safety from using an airbag helmet is large enough not to be outweighed by the loss in convenience or looks or whatever. This seems pretty difficult to me, especially as I bet the convenience gain isn't very big: I bet an effective airbag helmet needs to fit you about as well as an ordinary helmet, and it'll need to be put on with about as much care.
hop 16y ago
The acceleration your car experiences in an accident is an order of magnitude higher than anything your car ever experiences in everyday use, making the control system very easy to engineer - something like anything over 20g's and it pops (the algorithms take vehicle speed into account too for a varying threshold and probably use jerk.) In that video where the rider was hit from behind, there was very low relative change in velocity of the sensors in that helmet - no direct impact spike - and its on par with the same levels of acceleration you would experience in everyday bike riding.
- varjag 16y ago
  No, it was not easy at all with the cars. You can easily have 20g impact when you speed over road bump. Also, I'm sure in the 5 years they spent on product R&D, the concerns that came up to you in few minutes here were raised too.
  - hop 16y ago
    You can easily have 20g impact when you speed over road bump. The accelerometers in your car are bolted to your 4,000lb frame, riding on dampened shocks and rubber tires - you will never experience acceleration anywhere close to what you would in a frame-on collision. They don't have a shipping product yet.
    - varjag 16y ago
      You will easily experience that, just that energies involved are minuscule compared to collision. The mass of your car frame is irrelevant, shockwaves still propagate through solid bodies. When someone slaps your face hard, that can easily be 100s of Gs without much ill effect.

forgottenpaswrd 16y ago

That's what I was thinking, e.g look at the impacts on mtb freestyle: http://www.youtube.com/watch?v=DDaqcEkQRnM&feature=related http://www.youtube.com/watch?v=DDaqcEkQRnM&feature=relat... You can use tracker http://www.cabrillo.edu/~dbrown/tracker/ http://www.cabrillo.edu/~dbrown/tracker/ to calculate the decelerations on jumps. They are huge, bigger than the car hitting the back on the bike. The only way to do this is to monitor the direction of the deceleration, not only the magnitude.