3 ms·

Much is made of the "self-driving" part of autonomous vehicles, but I think "self-parking" could end up being just as significant. Imagine if there was no longe

Much is made of the "self-driving" part of autonomous vehicles, but I think "self-parking" could end up being just as significant. Imagine if there was no longer any reason to mandate parking capacity for planning and zoning purposes; you drive to your destination, and your car goes and parks itself (paying/negotiating automatically if needed), and returns when summoned via phone.

majormajor 7y ago
This sounds like high amounts of traffic: a bunch of people being dropped off and being picked up at arbitrary locations with the cars then having to go to and come back from various other bottlenecked hotspots.
- zaroth 7y ago
  I wouldn’t jump to that conclusion right away. You open up a lot of free space if you can take away street parking, which could also significantly alleviate traffic, for example with dedicated lanes for small-personal battery powered bikes, scooters, etc. It’s also not necessary for vehicles to travel several miles to find parking. Typical urban parking is designed so the average person is walking maybe a couple hundred feet maximum? That’s a lot of localized parking. If the car was traveling just a few thousand feet on average to find a lot, that’s not a lot of extra parking-miles traveled, but still a paradigm shift in parking process. Also, you eliminate the issue of people circling the block trying to find parking, which is adding slow traffic in dense areas. Finally, it’s likely the vehicle is not going to park, but rather off to pick up its next ride. At the most optimized, there are pickup and drop off points at every block where someone gets out, and immediately someone else gets in, and off the vehicle goes to the next stop.
  - majormajor 7y ago
    > Finally, it’s likely the vehicle is not going to park, but rather off to pick up its next ride. Yeah, a managed fleet system could have very different congestion characteristics, but the proposal was specifically that self-parking might be a game changer on its own. At that point it's basically valet parking, and you still have to stash the cars somewhere, and they still have to go back and forth between that somewhere and wherever the passengers got out. You get rid of the circling but I don't think you get rid of queuing and grid backups.
    - lovemenot 7y ago
      There's a lot of automatic parking in Tokyo and other cities in Japan. Drive up to the building, which has space for just one car in the "lobby". Get out of the car, enter data into the console. Your car then enters the building and is stored by a robotic system. Later, do the same in reverse.
  - jcranmer 7y ago
    > Typical urban parking is designed so the average person is walking maybe a couple hundred feet maximum? That’s a lot of localized parking. Typical urban parking outside of car-centric US cities is "your car takes up as much space as your apartment/hotel room/office cubicle, where the hell do we have space to put it?" For someplace as crammed as Manhattan, the parking isn't "around the block", it's on the other side of the traffic choke points. > Finally, it’s likely the vehicle is not going to park, but rather off to pick up its next ride. I don't buy that. Traffic patterns in large cities end up showing a distinct bimodal pattern: there is a large influx of people into the central business district in the morning, and then a large outflux of people in the afternoon. If I'm commuting by self-driving car, and I get in at 8:45-ish, there's going to be no demand for my car since everyone who needs to be in work by 9:00 is on the way. You can see this in systems that are already optimized to move people en masse to/from cities: commuter rail and bus systems. Large city railroad stations often have rail yards for train storage near their downtown hubs simply to store the trains during the day.
    - Karrot_Kream 7y ago
      You can also see this with bike and scooter shares. Bikes and scooters clump up at hubs that enter the transit system in the morning and empty out in the evening commute.
- notatoad 7y ago
  It would be a lot of traffic, but circling the block to find a parking spot is already such a huge contributor to traffic that it might still be a net win

majormajor 7y ago

This sounds like high amounts of traffic: a bunch of people being dropped off and being picked up at arbitrary locations with the cars then having to go to and come back from various other bottlenecked hotspots.

zaroth 7y ago
I wouldn’t jump to that conclusion right away. You open up a lot of free space if you can take away street parking, which could also significantly alleviate traffic, for example with dedicated lanes for small-personal battery powered bikes, scooters, etc. It’s also not necessary for vehicles to travel several miles to find parking. Typical urban parking is designed so the average person is walking maybe a couple hundred feet maximum? That’s a lot of localized parking. If the car was traveling just a few thousand feet on average to find a lot, that’s not a lot of extra parking-miles traveled, but still a paradigm shift in parking process. Also, you eliminate the issue of people circling the block trying to find parking, which is adding slow traffic in dense areas. Finally, it’s likely the vehicle is not going to park, but rather off to pick up its next ride. At the most optimized, there are pickup and drop off points at every block where someone gets out, and immediately someone else gets in, and off the vehicle goes to the next stop.
- majormajor 7y ago
  > Finally, it’s likely the vehicle is not going to park, but rather off to pick up its next ride. Yeah, a managed fleet system could have very different congestion characteristics, but the proposal was specifically that self-parking might be a game changer on its own. At that point it's basically valet parking, and you still have to stash the cars somewhere, and they still have to go back and forth between that somewhere and wherever the passengers got out. You get rid of the circling but I don't think you get rid of queuing and grid backups.
  - lovemenot 7y ago
    There's a lot of automatic parking in Tokyo and other cities in Japan. Drive up to the building, which has space for just one car in the "lobby". Get out of the car, enter data into the console. Your car then enters the building and is stored by a robotic system. Later, do the same in reverse.
- jcranmer 7y ago
  > Typical urban parking is designed so the average person is walking maybe a couple hundred feet maximum? That’s a lot of localized parking. Typical urban parking outside of car-centric US cities is "your car takes up as much space as your apartment/hotel room/office cubicle, where the hell do we have space to put it?" For someplace as crammed as Manhattan, the parking isn't "around the block", it's on the other side of the traffic choke points. > Finally, it’s likely the vehicle is not going to park, but rather off to pick up its next ride. I don't buy that. Traffic patterns in large cities end up showing a distinct bimodal pattern: there is a large influx of people into the central business district in the morning, and then a large outflux of people in the afternoon. If I'm commuting by self-driving car, and I get in at 8:45-ish, there's going to be no demand for my car since everyone who needs to be in work by 9:00 is on the way. You can see this in systems that are already optimized to move people en masse to/from cities: commuter rail and bus systems. Large city railroad stations often have rail yards for train storage near their downtown hubs simply to store the trains during the day.
  - Karrot_Kream 7y ago
    You can also see this with bike and scooter shares. Bikes and scooters clump up at hubs that enter the transit system in the morning and empty out in the evening commute.
notatoad 7y ago
It would be a lot of traffic, but circling the block to find a parking spot is already such a huge contributor to traffic that it might still be a net win