Tesla Auto park : parallel-parking - Why can't the car also pull out of the spot Mercedes

[Bebop]

It works for both parallel and perpendicular on both the MS & MX.

Edit: Noticed you have HW2.x, so I'm not sure if that's been enabled yet on post-AP1 cars.

Ah interesting. Well I do have HW2.5 so I’m wondering if it has to be enabled at that point.

I’ve seen the P logo, but while I’m on local roads in traffic.

 

[JohnnyG]

Ah interesting. Well I do have HW2.5 so I’m wondering if it has to be enabled at that point.

I’ve seen the P logo, but while I’m on local roads in traffic.

Okay, so be honest... have you ever been curious, even if just a little, to engage autopark when that pops-up and see what happens?? I know I've thought about it!! lol

On a more serious note though, there's some good information on page 78 of my owner's manual. Your page may vary, but it's under the Autopark subheading, under the Driver Assistance section. Here's a couple notes of interest...

1. Autopark detects perpendicular parking locations when driving below 10 mph (16 km/h).
2. Autopark detects potential perpendicular parking spaces that are at least 9.5 feet (2.9 meters) wide with a vehicle parked on each side.
3. If the Autopark icon does not appear at potential parking spaces when driving at the indicated speed, it is possible that Autopark is calibrating. Autopark requires a calibration process when Model S is new, or when tires are changed.
   1. During a parking sequence, Autopark must maneuver Model S with a great deal of precision. Therefore, before it can be used, Autopark must complete a calibration process. Calibration can take anywhere from 30 minutes to several days, depending on driving behavior. When Autopark is calibrating, a note displays on the Driver Assistance settings screen indicating that calibration is in progress. When calibration is complete, this note no longer displays and Autopark is available for use.
   2. Note: Autopark repeats the calibration process whenever tires are changed.
4. Autopark is particularly unlikely to operate as intended in the these situations.
   1. The road is sloped. Autopark is designed to operate on at roads only.
   2. The curb is constructed of material other than stone, or the curb cannot be detected.

Someone else with HW2.5 might be able to confirm whether they're having issues with it too?

 

[BrettS]

Okay, so be honest... have you ever been curious, even if just a little, to engage autopark when that pops-up and see what happens?? I know I've thought about it!! lol

It would attempt to park in the space that it thinks exists. It uses the ultrasonic sensors as it parks and would almost certainly determine that there is now a car in the space it believed to be empty and abort the parking attempt. I’m not sure it would really be too exciting although I imagine it would really confuse the people around you on the road.

 

[RogerHScott]

It would attempt to park in the space that it thinks exists. It uses the ultrasonic sensors as it parks and would almost certainly determine that there is now a car in the space it believed to be empty and abort the parking attempt. I’m not sure it would really be too exciting although I imagine it would really confuse the people around you on the road.

That's the sort of "obvious" situation -- new obstacle where there wasn't (perceived) to be one when the P appeared. More interesting, perhaps, is what happens if the "guide" cars it must have thought existed on either side, and/or the curb, "disappear" mid-parking? Or if you really want to mess with its head, what if you just move one or the other adjacent cars forward or backward a little during the procedure?

 

[BrettS]

That's the sort of "obvious" situation -- new obstacle where there wasn't (perceived) to be one when the P appeared. More interesting, perhaps, is what happens if the "guide" cars it must have thought existed on either side, and/or the curb, "disappear" mid-parking? Or if you really want to mess with its head, what if you just move one or the other adjacent cars forward or backward a little during the procedure?

It might be interesting to play around with and see what happens, but I suspect it wouldn’t really matter. I think it just uses the ‘guide’ cars to find a space that’s big enough to park in and once that happens it’s done with them. Obviously the goal is to avoid hitting the cars around it while it’s parking, so it attempts to plot a path that will avoid the cars. If the cars disappear it shouldn’t matter as I don’t know that it actually uses them as guides during the parking maneuvers.

This is just speculation, of course, so it might still be interesting to try at some point just to see what happens.

 

[JohnnyG]

That's the sort of "obvious" situation -- new obstacle where there wasn't (perceived) to be one when the P appeared. More interesting, perhaps, is what happens if the "guide" cars it must have thought existed on either side, and/or the curb, "disappear" mid-parking? Or if you really want to mess with its head, what if you just move one or the other adjacent cars forward or backward a little during the procedure?

Okay, I like where you're going with this! Now we need volunteers to try these scenarios. If someone writes-up a test script with all the scenarios they are curious about (within reason...), I'm more than willing to go attempt them for the sake of analytic research!
To be clear, not anything that would endanger others or myself.

 

[RogerHScott]

It might be interesting to play around with and see what happens, but I suspect it wouldn’t really matter. I think it just uses the ‘guide’ cars to find a space that’s big enough to park in and once that happens it’s done with them. Obviously the goal is to avoid hitting the cars around it while it’s parking, so it attempts to plot a path that will avoid the cars. If the cars disappear it shouldn’t matter as I don’t know that it actually uses them as guides during the parking maneuvers.

This is just speculation, of course, so it might still be interesting to try at some point just to see what happens.

I think the adjacent cars matter more than that, because it is explicitly trying to park between them, which is why it requires them to be there in the first place. If one of them moves into its path, obviously it will be detected like any other obstacle and likely just result in the entire parking maneuver aborting. But I'm curious about what happens if one moves away, such that the "between" location changes mid-maneuver.

 

[BrettS]

I think the adjacent cars matter more than that, because it is explicitly trying to park between them, which is why it requires them to be there in the first place. If one of them moves into its path, obviously it will be detected like any other obstacle and likely just result in the entire parking maneuver aborting. But I'm curious about what happens if one moves away, such that the "between" location changes mid-maneuver.

I dunno. I feel like it’s already determined the size of the space before it starts parking, so it should be able to determine where it is and where it should end up without using those other cars, but I don’t know for sure. I guess we just need @JohnnyG to test for us

 

[MarcusMaximus]

As I remember, the parallel parking maximum speed for autopark to show up is 15mph. I seem to remember it being much lower, single digits, for perpendicular parking. I generally find perpendicular autopark to be useless, so didn't pay much attention.

 

[RogerHScott]

As I remember, the parallel parking maximum speed for autopark to show up is 15mph. I seem to remember it being much lower, single digits, for perpendicular parking. I generally find perpendicular autopark to be useless, so didn't pay much attention.

Although the value of parallel autopark is greater, the (practical) opportunities to use it are far fewer, in my experience. So, overall, I think their net utilities are about even, if not with perpendicular a bit ahead.