Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Yes!!! Enhanced summon is on its way.....

This site may earn commission on affiliate links.
A lot of "memory" and redundancy might come from the maps that AP uses, currently and in the future. No reason the map couldn't specify which traffic light to pay attention to when you are in a given lane. And while the car itself doesn't learn, it's easy for it to confirm the route went as expected with the lanes specified or the driver took over an did something else. Then the navigation server and maps can be updated as necessary to get the desired outcome. No change in car firmware.
 
  • Like
Reactions: bedoig
A lot of "memory" and redundancy might come from the maps that AP uses, currently and in the future. No reason the map couldn't specify which traffic light to pay attention to when you are in a given lane. And while the car itself doesn't learn, it's easy for it to confirm the route went as expected with the lanes specified or the driver took over an did something else. Then the navigation server and maps can be updated as necessary to get the desired outcome. No change in car firmware.

I think traffic (congestion) data available to Tesla is not lane-specific. In that case, it may not help with the lane changes.
 
Enhanced Summon and FSD are going to merge one day.

It might as well be now - in this conversation.

With the new AP/FSD designations, the lines are now blurred. NoA and Advanced Summon (AS) used to be part of EAP but now are part of FSD. Both features were supposed to not need HW3, and now there's speculation that they will now use HW3.

Assuming FSD people will get HW3, here's the breakdown:
1) EAP only: No HW3. Get NoA, AS.
2) EAP+FSD, AP+FSD: Get HW3, ENoA? EAS? (Enhanced versions?)

It seems to me that NoA's current issues aren't really "vision" problems but more algorithmic. It can already detect cars on the highway, it just isn't smart enough to do things well. So I'm not sure what HW3 would really add. But let's say I'm wrong and ENoA makes things better. NoA has to have some minimal level of safety and reliability. What kind of things could HW3/ENoA do over NoA?

Since AS is dealing with a more dynamic environment (other cars, pedestrians, stop signs, shopping carts, etc) I could see it benefiting from additional processing hardware.
 
I do think Advanced Summon will be an improvement over the current Summon. Right now, current summon just moves your car forward or backwards a little bit. Sure, that's nice if you need to move your car out of a tight spot but that's it. Hopefully, with advanced summon, it will be a bit more useful since the car will be able to also steer left and right to come to you. So I could see a situation where you are parked near the entrance of a store and it is pouring rain. Advanced Summon might work to bring the car closer to you which the old summon would not be able to do. But I have no illusions. This first version of Advanced Summon will not be "drop you off and go find a parking lot on its own". However, I do think it lays the groundwork to eventually become that in the future.
I can see into the future, raining, some South Bay watering hole, a line of customers all peering out into the storm, patiently waiting on the log jam of Tesla’s coming to collect their owners...hehehe. Not any time soon.
 
I can see into the future, raining, some South Bay watering hole, a line of customers all peering out into the storm, patiently waiting on the log jam of Tesla’s coming to collect their owners...hehehe. Not any time soon.

...and those lucky few with FSF getting away the fastest. Full self flying with the HW8.0 e-powered air jets for low level flight.
 
My biggest grip with NoA is how early it asks me to change lanes, like 7 miles before I exit or change highways. At times it wants to put me right in the center of completely stopped lanes. I drive all the NorCal freeways, 80, 680, 101, 580, 20, 237, 50, you name it and the NoA is currently not active on my trips. First world problem I know, but hope it becomes smarter as time goes on.
 
My biggest grip with NoA is how early it asks me to change lanes, like 7 miles before I exit or change highways. At times it wants to put me right in the center of completely stopped lanes. I drive all the NorCal freeways, 80, 680, 101, 580, 20, 237, 50, you name it and the NoA is currently not active on my trips. First world problem I know, but hope it becomes smarter as time goes on.
And if they change that the opposite complaint is that NoA can’t successfully complete an exit because it starts trying too late. It’s a hard balance, especially since Tesla is unwilling to use their public fleet to experiment with aggressive lane changes
 
I would not jump right to the boss fight when I get this feature.

First, I'll see if I can summon it out of the garage. If it can handle that, I'll put the car back in the garage and see if I can summon it to the end of my driveway. If that's good, I'll walk down my street to a point where I can keep eyes on the car and see if it can find me. Next, I'll put a big cardboard box in the driveway and see what the car does. It will be fun designing some levels leading up to the boss fight.

But for now, I won't be attempting to summon it at the grocery store the day before a snow storm.

What I also find interesting is that there are no lane markers in parking lots, which means IF summon is in decent enough shape to release, the car's ability to drive on unmarked roads might be improved as well. Curious to test this scenario too. I have a few roads around me that start with lane markings and then transition to no markings, so I have a good idea at how well existing AP does (not great).

I like your cautious approach and also look forward to testing it under safe circumstances. It’s going to be fun if tested safely.
 

Wow. That matches EXACTLY with what the deleted thread reported which was that it moves incredibly slowly like a confused child.

I also like how they tested it in a completely empty parking lot. Kind of like oggling a photo of a gorgeous playmate and asking what is the last book she read.

Paging @Garlan Garner so we can hear about how excited he still is about this.
 
Wow. That matches EXACTLY with what the deleted thread reported which was that it moves incredibly slowly like a confused child.

Remember that Auto Steer was limited to slow speeds at first too until Tesla got more confident in the system. Once Enhanced Summon gets better, I am confident Tesla will raise the max speed. Considering that this is the first feature designed to operate with no driver in the car, I don't blame Tesla if they are extra conservative.

Well that looks ... utterly useless.

Yes but keep in mind, it's a first iteration of the feature. It will get better over time. The important thing to me is that this is the first feature designed to operate with no driver. And the car can navigate and steer with no driver. As the feature gets more confident, it will become more useful.
 
Remember that Auto Steer was limited to slow speeds at first too until Tesla got more confident in the system. Once Enhanced Summon gets better, I am confident Tesla will raise the max speed. Considering that this is the first feature designed to operate with no driver in the car, I don't blame Tesla if they are extra conservative.



Yes but keep in mind, it's a first iteration of the feature. It will get better over time. The important thing to me is that this is the first feature designed to operate with no driver. And the car can navigate and steer with no driver. As the feature gets more confident, it will become more useful.

It is your prediction that it will get more useful. Maybe it will, maybe it won't. Maybe it will be useful only with HW3 upgrade, or HW4, or inclusion of newer sensors, etc.

My view is this is another case of Elon overpromising and epically under-delivering. Remember, HW 2.5 was going to be good enough for the car to be an autonomous taxi. Based on what we see in this video how many years before that is even possible?

Musk is brilliant but he strikes me on too many occasions as the crappy manager who thinks something is possible and keeps firing the engineers telling him it's not till he finds some intern who says he can do it.

Then you get "enhanced" summon that looks like it is going to be worthless outside of use on empty private property.
 
It is your prediction that it will get more useful. Maybe it will, maybe it won't. Maybe it will be useful only with HW3 upgrade, or HW4, or inclusion of newer sensors, etc.

My view is this is another case of Elon overpromising and epically under-delivering. Remember, HW 2.5 was going to be good enough for the car to be an autonomous taxi. Based on what we see in this video how many years before that is even possible?

Musk is brilliant but he strikes me on too many occasions as the crappy manager who thinks something is possible and keeps firing the engineers telling him it's not till he finds some intern who says he can do it.

Then you get "enhanced" summon that looks like it is going to be worthless outside of use on empty private property.

This is just a basic version that Tesla knows is safe for public release and then as the system improves, they will improve it. It's an iterative process. Remember how useless auto steer was when it was first released and look how good it is now, 2.5 years later?
 
Except for being too slow, it's pretty good for what I was expecting the first version of Enhanced Summon to be: "you summon the car to a designated spot". Over time, it will get better.

We are just seeing the first version that will eventually become this: "Summon: your parked car will come find you anywhere in a parking lot. Really."