Experts ONLY: comment on whether vision+forward radar is sufficient for Level 5?

[stopcrazypp]

Its over. its done.
Its 7.6 miles, deal with it. About the same miles/stops others have traveled while doing cross country.


“It will do this without the need for a single touch, including the charger,” says Musk.

Elon directly contradicts you. unless in this case Elon statement doesn't amount to anything cause it doesn't fit your agenda? That won't be surprising would it? you have a knock at attempting to twist things to look your way.

But we both know that we will know very easily the route that tesla takes by looking at which supercharger is automated.

its over man, you were wrong. let it go...

Lol, you were wrong about 3-5 miles tops, even using your cherry picked stations. This is putting aside the point that is already more than the ~2 miles you suggested earlier before factoring in charging stops. I just put examples from the default routes from two charging planning apps (including the one you used), and they involved a lot more off highway travel than you let on. And even in the article you linked, it suggests to use more stops rather than skipping. Skipping actually takes a lot longer.

As for the snake chargers, I did not initially focus on that part of his trip promise. Although he says the driver does not have to touch the charger, he did not say it necessarily has to be snake chargers or some kind of robotic solution. Tesla had supercharger valets since a year back at busy locations. They can have similar for the trip if it comes to it. This would be similar to some states that require an attendant at gas pumps.
Tesla Supercharger Valet Service In Action - Video

 

[Bladerskb]

5 and 7 miles is practically the same thing and valet service? now you are just reaching...

 

[bob_p]

Because the coast-to-coast trip will use the automatic connector, Tesla will likely pre-plan the route, select the superchargers that will be used, and upgrade a charging spot to have the new automatic connector.

When the demo is run, they'll likely have an advance team that will run ahead of the demo car - block out the automatic connector spot, and make sure it is operational and available before the demo car gets there.

And it's likely they will have tested the route, either completely or in segments before doing the actual demo - to ensure they really can go coast-to-coast without human involvement in driving or charging.

If they can achieve this, even with the preparations, it will be a significant demonstration of the capabilities Tesla vehicles could have at some point (and hopefully soon).

 

[Reciprocity]

I dont know why there is any focus on the car charging itself. The Nov-Dev demo is not a self charging demo, its a self driving demo. Tesla doesn't need automated charging until they have FSD in the market in a major way. Then they will need a solution because cars in the Tesla network will require charging throughout the day while their owners are at work.

 

[SabrToothSqrl]

Professional Internet Reader here.

Seems full autonomy is fully functional today using only two 'cameras'... so... yes?

 

[JeffK]

I dont know why there is any focus on the car charging itself. The Nov-Dev demo is not a self charging demo, its a self driving demo. Tesla doesn't need automated charging until they have FSD in the market in a major way. Then they will need a solution because cars in the Tesla network will require charging throughout the day while their owners are at work.

I don't know, here's a quote:

I feel pretty good about this goal is that we will be able to demonstrate a demonstrtion drive of our full autonomy all the way from LA to New York. So basically from home in LA to let’s say dropping you off in Times Square, NY and then having the car parking itself by the end of next year (2017) without the need for a single touch including the charger.

 

[Reciprocity]

I don't know, here's a quote:

It silly for him to have said and its silly for us to by trying to figure out how many miles the car will have to drive between interstate and charging stations. In theory, they could just have valets and then the driver doesn't have to touch the charging. I think most will be focused on the driving part and not the charging part.

 

[iqless]

are we allowed to ask questions on this thread if we're not qualified

i'm too scared to even talk

 

[SabrToothSqrl]

are we allowed to ask questions on this thread if we're not qualified

i'm too scared to even talk

Who said that?! Unauthorized activity in sector 7G.

 

[lunitiks]

Sorry guys, late to this thread.

Maybe.

 

[MarcusMaximus]

It silly for him to have said and its silly for us to by trying to figure out how many miles the car will have to drive between interstate and charging stations. In theory, they could just have valets and then the driver doesn't have to touch the charging. I think most will be focused on the driving part and not the charging part.

This whole thing is a pretty silly conversation. I'm not even sure what we're arguing about anymore. Why on earth are we trying to read the tea leaves of what Musk said about a hypothetical route that may or may not be taken in Nov. or Dec. and what that means about specific dates for when FSD is going to be ready? This thread has gone totally off the rails.

 

[Reciprocity]

This whole thing is a pretty silly conversation. I'm not even sure what we're arguing about anymore. Why on earth are we trying to read the tea leaves of what Musk said about a hypothetical route that may or may not be taken in Nov. or Dec. and what that means about specific dates for when FSD is going to be ready? This thread has gone totally off the rails.

No doubt, but it is fun to speculate. It gets less fun when people over personalize it. Everyone is seeing a lot of the same content and coming to different conclusions and that's fine. I learn a lot here about things I had no idea about and I try to incorporate that into my opinion. I chime in when I see something that I find to be not accurate based on the facts as I know them, but there can always be other facts that I dont know about. I try to read as much source material that is presented here and elsewhere as possible.

 

[stopcrazypp]

It silly for him to have said and its silly for us to by trying to figure out how many miles the car will have to drive between interstate and charging stations. In theory, they could just have valets and then the driver doesn't have to touch the charging. I think most will be focused on the driving part and not the charging part.

That's kind of the thing I was getting at. Having a robotic charger at every location is simply not important (personally, I didn't even notice that part in the announcement) nor did Elon said that they would have robotic chargers, only that the driver didn't have to make a single touch.

 

[Kant.Ing]

An FSD demo does not mean real FSD, as the software teams will be busy turning in their annual focal review.

Just an old idea

 

[Bladerskb]

Do you know what it will mean when Tesla demonstrates highway autonomy? It will mean they have a huge lead on everyone else. They will have demoed coast to coast full autonomy and have over 100,000 cars equipped and more then likely activated as a safety feature until it can be approved by regulators who will be falling over each other to get it done to take credit for saving thousands of lives. And where will Intel be? Same place they are today, paying 15 billion for the chip that can't see a giant semi truck passing across it's face.

You do realize that the eyeq3 wasn't designed for lateral car detection and that mobileye actually have Eyeq4 with software ready for self driving car? Yet tesla can't even match 10% of eyeq3 features. It can't even read signs, traffic light, pavement markings, differentiate cars from trunks, motorcycle and pedestrians and see cars in other lane for starters, see potholes, bumps, trash, road edge. There's a lot more features from eyeq3, current AP pales in comparison.

Exclusive: The Tesla AutoPilot - An In-Depth Look At The Technology Behind the Engineering Marvel - Page 4 of 6

Tesla simply recreated the lane identifying and following part of eyeq3.
It still struggles heavily even with the latest update.

almost killed the driver
Untitled - Streamable

Local roads

 

[croman]

You do realize that the eyeq3 wasn't designed for lateral car detection and that mobileye actually have Eyeq4 with software ready for self driving car? Yet tesla can't even match 10% of eyeq3 features. It can't even read signs, traffic light, pavement markings, differentiate cars from trunks, motorcycle and pedestrians and see cars in other lane for starters, see potholes, bumps, trash, road edge. There's a lot more features from eyeq3, current AP pales in comparison.

Tesla simply recreated the lane identifying and following part of eyeq3.
It still struggles heavily even with the latest update.

almost killed the driver
Untitled - Streamable

Local roads

What other car manufacturers have cars available to purchase that do the things AP1 does with the EyeQ3? Why does ME get all the credit but Tesla only gets scorn? It can't be only ME if Tesla AP1 is the only iteration of Eyeq3 that does anything worthwhile.

 

[Bladerskb]

What other car manufacturers have cars available to purchase that do the things AP1 does with the EyeQ3? Why does ME get all the credit but Tesla only gets scorn? It can't be only ME if Tesla AP1 is the only iteration of Eyeq3 that does anything worthwhile.

2017 Volvo Pilot Assist 2 (It even handles curves better than AP1 and AP2). There's no dumb braking and not recognizing car infront, etc... The reason you are unaware is because those other car manufacturers don't have thousands of fanatics posting daily videos about their lv 2 systems. Obviously Volvo doesn't have those fanatics so there are no thousands of video of their pilot assist 2 randomly driving around.

rain/snow/sleet

local road

huge bend

 

[JeffK]

It can't even read signs, traffic light, pavement markings, differentiate cars from trunks, motorcycle and pedestrians and see cars in other lane for starters, see potholes, bumps, trash, road edge. There's a lot more features from eyeq3, current AP pales in comparison.

The Tesla self driving demo proves this assumption incorrect as these items are clearly identified in the right hand videos.

You're simply commenting on the quickly built EAP implementation. MobilEye has been at it for 18 years and Tesla's implementation was just built in a few months.

You are also posting videos of people not using it on proper highways whereas the Volvo videos you posted are either on highways or with a clearly marked solid double yellow line.

No one is claiming the current EAP is on par with AP 1.0 yet it's a work in progress. As far as EyeQ3 implementations, it was widely recognized that Tesla had the best implementation in the industry.

 

[Bladerskb]

The Tesla self driving demo proves this assumption incorrect as these items are clearly identified in the right hand videos.

You're simply commenting on the quickly built EAP implementation. MobilEye has been at it for 18 years and Tesla's implementation was just built in a few months.

You are also posting videos of people not using it on proper highways whereas the Volvo videos you posted are either on highways or with a clearly marked solid double yellow line.

No one is claiming the current EAP is on par with AP 1.0 yet it's a work in progress. As far as EyeQ3 implementations, it was widely recognized that Tesla had the best implementation in the industry.

I'm not talking about some demo that was based on nvidia driveworks. I'm talking about the current capabilities of ap2, for example it can t change speed based on speed limit and it clearly can t see and doesn't take into account road edges or different config of vehicles nor pedestrians or cars on multiple lanes

I posted you local roads with volvo pilot assist. Secondly ap2 fail even with GREAT lane markings as you can see in the videos of the latest update.

Volvo pilot assist 2 is car better than ap2 and probably even better than ap1.

All those false positive/negative brake issues, ramming into the car ahead cause it didn't see it, truck lust, etc. Doesn't happen. With volvo pilot assist 2.

I can build a better implementation of ap2 in a few days. In fact I don't have to, geohut already has. His openpilot 0.3 is better than ap2. That's just one guy in a couple weeks. You bringing up Mobileye and saying 18 years shows how naive you are. The technology and the scientific breakthrough 18 years ago pales in comparison to today. I can do something in 1 hour that it takes me a decade to do 18 years old.

Tesla having the best eyeq3 implementation in the industry is another myth

 

[Bladerskb]

Another myth is tesla's supposed lead in tech and horde of IP.
But when you actually take a look it, tesla has the least amount of patents among car makers, infact. Other car makers patent more tech a year than tesla has in total.

Mobileye for example has a horde of patents relating to driverless cars and vision. you think they won't come for tesla when the time is right?

Patents by Assignee Mobileye Vision Technologies Ltd. - Justia Patents Search

here is mobileye 8 camera system that tesla copied. including the exact placement.
36mins 24secs