3 ms·

A nice engineering project, sure. But it has nothing to do with cars. They only thing that their results show is that you can get vastly different results for v

by adwn 12y ago

A nice engineering project, sure. But it has nothing to do with cars. They only thing that their results show is that you can get vastly different results for vastly different constraints and requirements.

onion2k 12y ago
It does have something to do with cars - it shows that if we changed our requirements we could be driving much more efficient cars. Even if safety and comfort meant than we could only get 1/10th of the demonstration, that'd still be a good deal better than what we have now.
- jerf 12y ago
  You need to drop a lot of the safety, grossly raise the price, or both though. I've got some family in the auto industry, and their consensus is that we are either really rapidly coming up on the point where we've got just about all the safety and fuel economy we can have without sacrificing safety, or we're already there. You're getting into the space where arguing for increased fuel economy is going to be directly arguing for killing more people. You may find this a tough sell. Especially since telling the automakers to "make less safe cars" is also going to have to come with an ironclad guarantee that when people die in these cars, the car companies won't be sued. And getting that law passed is going to be a tough sell, too, for a lot of reasons. The safety advocates aren't exactly going to stand passively by while you try to tear their stuff apart. Basically, you can't have it all. Not even if you pass a law saying that you have to have it all. And while we're at it, let us also not forget that "grossly raise the price" also is not something you can just wave into being. The Democrats can hardly try to make "inequality" into a major issue, then do something that unambiguously makes cars much more expensive for everybody.
  - onion2k 12y ago
    I don't have any family with a vested interested so perhaps I can be more impartial. ;) There is a very straightforward way to make cars both more efficient and safer - make them smaller and lighter. Then they'll require less energy to move around and they'll have less energy if they hit things. The challenge to overcome is that driving a small and light car is actually more dangerous while big and heavy cars are still on the road. So we have to get rid of them. That's actually very easy too - ramp up taxes on fuel to the point where people actually give a damn. It's what we're doing here in Europe, and the proliferation of small, light cars is very noticeable.
    - sigzero 12y ago
      "smaller and lighter" means more deaths when not all the potential impact cars are not also "smaller and lighter". Also, "smaller and lighter" and certainly what they used in the test isn't FAMILY practical.
    - jerf 12y ago
      I don't have a vested interest either; I don't work in the industry myself. What I have is a bit of a view into just how far down the road we are on the engineering tradeoffs already. Small and light cars aren't a miracle cure anyhow. It's amazing how small the gains are from shrinking a car. Yes, there are gains, but we're not talking taking a 40mpg sedan (in routine manufacturing now, BTW) and turning it into an 80mpg sedan "just by shrinking it a bit". You might get it down to 50mpg, but, per other articles observing that "mpg" is the wrong way to view it and "gallons per mile" starts making more sense, a 40 to 50 mpg gain is not actually that large. No, the environment can't be saved just by "making cars smaller". You don't win anywhere near as much as you think, because of the way drag works (it's nonlinear in speed). And you can't just start taking panels off and stripping out airbags. From what I gather, having trimmed just about everything they can from the spare tire, they're on the verge of dropping it off entirely. That's the level of optimization we're down to. There isn't much left after that. The seats are what they are due to safety regulations. The airbags are there because of safety regulations. The bumpers are there due to safety regulations. The body panels are what they are due to safety regulations and the need to survive a certain number of years out in the elements. There's just not much freedom in car design anymore. My family doesn't include designers in them, but they relayed the designer's complaints to me that the constraints are becoming so strong that it's becoming very difficult to differentiate between cars anymore. If it feels like they all look the same lately, excepting perhaps the exact location or shape of the lights, it's because they do look the same; the combination of safety and efficiency regulations is rapidly converging down to one solution for a given form factor. Again... this is the sign of an optimization process reaching its Pareto frontier, and that you can't get big gains anymore without big sacrifices somewhere else. You can look out in the world and see that last bit of evidence yourself, without anyone in the automotive industry. It's the same systems optimization problem faced by any engineer, even programmers.
  - ForHackernews 12y ago
    If we have self-driving cars that are dramatically less likely to be involved in accidents, then we can start relaxing the safety requirements.
    - jerf 12y ago
      Yes, this truly is the phase change that will get us out of this hole. Self-driving cars could also practically use battery-switch-out technology and very rich rental models where we don't all have to own our own cars anymore, reducing the actual cost of manufacturing by making fewer cars, another big win that you simply can't optimize the current system into.

by adwn 12y ago

onion2k 12y ago

It does have something to do with cars - it shows that if we changed our requirements we could be driving much more efficient cars. Even if safety and comfort meant than we could only get 1/10th of the demonstration, that'd still be a good deal better than what we have now.

jerf 12y ago
You need to drop a lot of the safety, grossly raise the price, or both though. I've got some family in the auto industry, and their consensus is that we are either really rapidly coming up on the point where we've got just about all the safety and fuel economy we can have without sacrificing safety, or we're already there. You're getting into the space where arguing for increased fuel economy is going to be directly arguing for killing more people. You may find this a tough sell. Especially since telling the automakers to "make less safe cars" is also going to have to come with an ironclad guarantee that when people die in these cars, the car companies won't be sued. And getting that law passed is going to be a tough sell, too, for a lot of reasons. The safety advocates aren't exactly going to stand passively by while you try to tear their stuff apart. Basically, you can't have it all. Not even if you pass a law saying that you have to have it all. And while we're at it, let us also not forget that "grossly raise the price" also is not something you can just wave into being. The Democrats can hardly try to make "inequality" into a major issue, then do something that unambiguously makes cars much more expensive for everybody.
- onion2k 12y ago
  I don't have any family with a vested interested so perhaps I can be more impartial. ;) There is a very straightforward way to make cars both more efficient and safer - make them smaller and lighter. Then they'll require less energy to move around and they'll have less energy if they hit things. The challenge to overcome is that driving a small and light car is actually more dangerous while big and heavy cars are still on the road. So we have to get rid of them. That's actually very easy too - ramp up taxes on fuel to the point where people actually give a damn. It's what we're doing here in Europe, and the proliferation of small, light cars is very noticeable.
  - sigzero 12y ago
    "smaller and lighter" means more deaths when not all the potential impact cars are not also "smaller and lighter". Also, "smaller and lighter" and certainly what they used in the test isn't FAMILY practical.
  - jerf 12y ago
    I don't have a vested interest either; I don't work in the industry myself. What I have is a bit of a view into just how far down the road we are on the engineering tradeoffs already. Small and light cars aren't a miracle cure anyhow. It's amazing how small the gains are from shrinking a car. Yes, there are gains, but we're not talking taking a 40mpg sedan (in routine manufacturing now, BTW) and turning it into an 80mpg sedan "just by shrinking it a bit". You might get it down to 50mpg, but, per other articles observing that "mpg" is the wrong way to view it and "gallons per mile" starts making more sense, a 40 to 50 mpg gain is not actually that large. No, the environment can't be saved just by "making cars smaller". You don't win anywhere near as much as you think, because of the way drag works (it's nonlinear in speed). And you can't just start taking panels off and stripping out airbags. From what I gather, having trimmed just about everything they can from the spare tire, they're on the verge of dropping it off entirely. That's the level of optimization we're down to. There isn't much left after that. The seats are what they are due to safety regulations. The airbags are there because of safety regulations. The bumpers are there due to safety regulations. The body panels are what they are due to safety regulations and the need to survive a certain number of years out in the elements. There's just not much freedom in car design anymore. My family doesn't include designers in them, but they relayed the designer's complaints to me that the constraints are becoming so strong that it's becoming very difficult to differentiate between cars anymore. If it feels like they all look the same lately, excepting perhaps the exact location or shape of the lights, it's because they do look the same; the combination of safety and efficiency regulations is rapidly converging down to one solution for a given form factor. Again... this is the sign of an optimization process reaching its Pareto frontier, and that you can't get big gains anymore without big sacrifices somewhere else. You can look out in the world and see that last bit of evidence yourself, without anyone in the automotive industry. It's the same systems optimization problem faced by any engineer, even programmers.
- ForHackernews 12y ago
  If we have self-driving cars that are dramatically less likely to be involved in accidents, then we can start relaxing the safety requirements.
  - jerf 12y ago
    Yes, this truly is the phase change that will get us out of this hole. Self-driving cars could also practically use battery-switch-out technology and very rich rental models where we don't all have to own our own cars anymore, reducing the actual cost of manufacturing by making fewer cars, another big win that you simply can't optimize the current system into.