3 ms·

Autonomous fleet parking has all the 3D density advantages of robot parking structures[1], but at dumb concrete prices. There's no aisle because there's no "ran

Autonomous fleet parking has all the 3D density advantages of robot parking structures[1], but at dumb concrete prices. There's no aisle because there's no "random access" of a particular vehicle. No extra headroom (because they're identical) extra width (because they're precise), ventilation (because they're electric), or even lights. It's the best of both worlds. Where land is cheap, simply drive the cars along a space-filling curve in a parking lot. One possible path is a double-spiral. Picture these lines[2] as the curb (except there's no curb, or even painted lines). The spiral path is simple enough to calculate. One can rotate the whole spiral as viewed from above, growing and shrinking it and "wear leveling" across the entire paved surface. And the technique should generalize to any shape parking lot. I expect writhing double-headed masses of autonomous cars near urban centers soon. This design eliminates entirely any 3D architecture, at the small expense of some wasted space in the center determined by the turning radius[3]. Anyone know the most efficient shape would be in 3D (geometrically and structurally)? A helix is the obvious extrapolation, but I'm not convinced it's optimal. [1] https://en.wikipedia.org/wiki/Automated_parking_system https://en.wikipedia.org/wiki/Automated_parking_system [2] https://openclipart.org/detail/216988/double-linear-spiral https://openclipart.org/detail/216988/double-linear-spiral [3] edit: now that I think about it, you can park cars there too

thaumasiotes 9y ago
> [no] ventilation (because they're electric) This seems... iffy. For one thing, my computers run on electricity, and they need ventilation. For another, I think it makes sense to have it be possible for humans to enter the car storage. On a similar note, it's a much worse idea to have your car constantly drive through the parking lot than to park it in the parking lot.
- schiffern 9y ago
  >This seems... iffy. For one thing, ... At first I wrote "no oversized ventilation," but I shortened it during revision after thinking, "no, no-one would think I meant something so dumb." Thanks for the correction. :P Yes, there will be ventilation, emergency lighting, and access paths. Those contingency systems are still much cheaper and take up less volume than the corresponding full-time systems. >it's a much worse idea to have your car constantly drive through the parking lot than to park it in the parking lot You say that as though it's obvious, but why? Parking lots are $$$, and this ~quadruples the density. It's not like you're paying drivers, so the only cost is a tiny amount of electricity (a parking robot would certainly use more, since A) it's heavier and B) it moves faster). I also presume the vehicles will be stationary most of the time, allowing them to de-energize power electronics in the common case. Wear-and-tear at those speeds is negligible, and multiple "lanes" means that if a car breaks down the other cars simply drive around it. Heck, at those speeds the following vehicle could push a disabled car out of a bottleneck (mature shared-use vehicle designs will likely feature non-scratch bumpers).

thaumasiotes 9y ago

> [no] ventilation (because they're electric) This seems... iffy. For one thing, my computers run on electricity, and they need ventilation. For another, I think it makes sense to have it be possible for humans to enter the car storage. On a similar note, it's a much worse idea to have your car constantly drive through the parking lot than to park it in the parking lot.

schiffern 9y ago
>This seems... iffy. For one thing, ... At first I wrote "no oversized ventilation," but I shortened it during revision after thinking, "no, no-one would think I meant something so dumb." Thanks for the correction. :P Yes, there will be ventilation, emergency lighting, and access paths. Those contingency systems are still much cheaper and take up less volume than the corresponding full-time systems. >it's a much worse idea to have your car constantly drive through the parking lot than to park it in the parking lot You say that as though it's obvious, but why? Parking lots are $$$, and this ~quadruples the density. It's not like you're paying drivers, so the only cost is a tiny amount of electricity (a parking robot would certainly use more, since A) it's heavier and B) it moves faster). I also presume the vehicles will be stationary most of the time, allowing them to de-energize power electronics in the common case. Wear-and-tear at those speeds is negligible, and multiple "lanes" means that if a car breaks down the other cars simply drive around it. Heck, at those speeds the following vehicle could push a disabled car out of a bottleneck (mature shared-use vehicle designs will likely feature non-scratch bumpers).