AP1 vs. AP2

[Tam]

...The hype is discouraging...

I think "hype" is a correct word in Tesla but I am not discouraged by it.

When Tesla first started, very few would think there's a market for EV and there was even a "Tesla Death Watch."

Driverless in general is predicted for 2020 the earliest or by 2021 and by then, Ford won't release it to average public but to specialized industry such as ride-hailing services.

The hype here is: Tesla intends to do a driverless cross country in about 8 months and its goal is to get the feature ready for public in 2018 (but the process of releasing it might be held up by laws.)

I feel both Tesla's AP1 and now AP2 hardware/firmware are progressing reasonably well even when taking in accounts of failures.

Ok! The timeline might be missed just like many times in the history but I don't see that's a reason to be full of gloom and doom about it.

If you were New York Times reporter Broder who test drove Tesla Model S in 2013, you would say:

"I drove a state-of-the-art electric vehicle past a lot of gas stations. I wasn’t smiling."

I don't think most are a "Broder"! And that's a real happy hype!

 

[Bladerskb]

I think "hype" is a correct word in Tesla but I am not discouraged by it.

When Tesla first started, very few would think there's a market for EV and there was even a "Tesla Death Watch."

Driverless in general is predicted for 2020 the earliest or by 2021 and by then, Ford won't release it to average public but to specialized industry such as ride-hailing services.

The hype here is: Tesla intends to do a driverless cross country in about 8 months and its goal is to get the feature ready for public in 2018 (but the process of releasing it might be held up by laws.)

I feel both Tesla's AP1 and now AP2 hardware/firmware are progressing reasonably well even when taking in accounts of failures.

Ok! The timeline might be missed just like many times in the history but I don't see that's a reason to be full of gloom and doom about it.

If you were New York Times reporter Broder who test drove Tesla Model S in 2013, you would say:

"I drove a state-of-the-art electric vehicle past a lot of gas stations. I wasn’t smiling."

I don't think most are a "Broder"! And that's a real happy hype!

it gets even worse. driverless cross country has nothing to do with urban fully self driving car and has been done many times.

 

[Electroman]

I am among those who prefer to pay up good money (not cheap at all) and get some imperfect technology right now and watch it grow rather than wait for perfection down the line.

as they say, perfection is the enemy of the good..

it gets even worse. driverless cross country has nothing to do with urban fully self driving car and has been done many times.

Really? Now I am beginning think of the 'T' word. You have gone from subtle to outright exaggerations to now statements that are totally false.

Most of us have been here long enough to notice a pattern..

 

[chillaban]

Really? Now I am beginning think of the 'T' word. You have gone from subtle to outright exaggerations to now statements that are totally false.

What part of "Kia has ap1 parity" led you to think that?

It's just awfully convenient that the examples for cars on the market that have AP1 parity are nowhere close and frankly not even AP2 parity. And the things that are better than Tesla Autopilot are yet to be released.

We've been watching Mercedes and Audi demo hands-off L3+ systems since 2012. Yet somehow when they get in customer hands, it turns into a >=47mph, ping pong between lane lines for 6 seconds before beeping, cannot handle cars cutting in/out of your lane for crap. Everyone is entitled to their opinion that we are fools to be happy with AP1 and the AP2 journey which is already pretty functional for a from-scratch system. But if you want to tell us that systems on the market exist that perform as well as AP1, and regurgitate the competitions' press releases that so far have all been vaporware or huge disappointments, that's where all credibility is gone.

For the record, I welcome competition for Autopilot. L2 ADAS systems where you can genuinely let the car handle throttle and steering for minutes or sometimes hours with zero intervention is basically a one player game right now: Tesla. We all know without competition, companies like to rest on their laurels and take advantage of customers. I would love to see another system on the market come to fruition. But for the last 4 years, I've gone through 3 cars and test drove 20+ different models, including the K900 ironically. So far, AP1 has been the market leader bar none, and AP2 is already fairly close to AP1. I think Mercedes DISTRONIC distance regulation is a bit smoother than AP2 at times but DISTRONIC is far worse than AP2 at brake ramp when approaching stopped traffic at 80mph. The Audi ACC implementation feels largely identical to the BMW one, and both are downright disgraceful in how they handle rapidly slowing traffic or cut-in detection. And as far as steering assistance systems: Nobody's anywhere close to AP2 or AP1. (I don't really have any brand loyalty. I take pride in doing my research and personal testing and choosing the best option for me at any given time)

I'll be reevaluating this in another year or two. Wake me up if something actually happens sooner.

 

[Bladerskb]

as they say, perfection is the enemy of the good..



Really? Now I am beginning think of the 'T' word. You have gone from subtle to outright exaggerations to now statements that are totally false.

Most of us have been here long enough to notice a pattern..

why don't you enter in google maps the direction from tesla HQ to time square and then see for yourself how many surface street miles there are. Let me help you out, a drive from Tesla HQ to Time Square has just 2 miles of surface street so that trip elon has been hyping not only has been done multiply times before but its also completely worthless as a fully self driving indicator. 0.0006% of the drive only matters to self driving. That's 2 miles out of 2,943.

 

[ABVA]

Autopilot hardware AP1 with MobilEye EyeQ3 were installed since 2014. People paid for it but its autopilot function was not activated until a whole year later. -------
And that's why Tesla is planning on a demonstration of driverless cross country LAX to NYC trip this December.

Amen to that

 

[Bladerskb]

What part of "Kia has ap1 parity" led you to think that?

It's just awfully convenient that the examples for cars on the market that have AP1 parity are nowhere close and frankly not even AP2 parity. And the things that are better than Tesla Autopilot are yet to be released.

We've been watching Mercedes and Audi demo hands-off L3+ systems since 2012. Yet somehow when they get in customer hands, it turns into a >=47mph, ping pong between lane lines for 6 seconds before beeping, cannot handle cars cutting in/out of your lane for crap. Everyone is entitled to their opinion that we are fools to be happy with AP1 and the AP2 journey which is already pretty functional for a from-scratch system. But if you want to tell us that systems on the market exist that perform as well as AP1, and regurgitate the competitions' press releases that so far have all been vaporware or huge disappointments, that's where all credibility is gone.

For the record, I welcome competition for Autopilot. L2 ADAS systems where you can genuinely let the car handle throttle and steering for minutes or sometimes hours with zero intervention is basically a one player game right now: Tesla. We all know without competition, companies like to rest on their laurels and take advantage of customers. I would love to see another system on the market come to fruition. But for the last 4 years, I've gone through 3 cars and test drove 20+ different models, including the K900 ironically. So far, AP1 has been the market leader bar none, and AP2 is already fairly close to AP1. I think Mercedes DISTRONIC distance regulation is a bit smoother than AP2 at times but DISTRONIC is far worse than AP2 at brake ramp when approaching stopped traffic at 80mph. The Audi ACC implementation feels largely identical to the BMW one, and both are downright disgraceful in how they handle rapidly slowing traffic or cut-in detection. And as far as steering assistance systems: Nobody's anywhere close to AP2 or AP1. (I don't really have any brand loyalty. I take pride in doing my research and personal testing and choosing the best option for me at any given time)

I'll be reevaluating this in another year or two. Wake me up if something actually happens sooner.

That's because the technology needed isn't ready for mass production till late 2017/early 2018, which is LIDAR and mobileye eyeq4.
but L3 cars will start showing up in 2018 with "Audi" and Nissan (both using eyeq4).

Meanwhile hands-free L2 will be here later this year in supercruise and ap1 parity will be here in Mercedes drivepilot 4.5 later this year too.

 

[TSLA Pilot]

We have a '15 MS with AP 1.0 and a '17 with AP 2.0.

Our '15 is much more sorted out and mature; 2.0 remains borderline dangerous, at least as of the mid-April, 2017 updates. The display constantly shows the lanes shifting wildly, even though they are not. Stability in corners is "iffy."

It's a work in progress.

If you want a decent AP, and that's your primary reason for buying the car, I'd buy a CPO with AP 1.0 for now.

Having said that, the AP video for 2.0 is pretty impressive. When we'll see it is the question that no one has answered just yet . . . .

 

[Bladerskb]

once eyeQ4 goes into mass production the flood gates will open up.

 

[GHammer]

Yea, they've said that before. The near future in this slide refers to the Q3. We know how that turned out.

 

[Bladerskb]

Yea, they've said that before. The near future in this slide refers to the Q3. We know how that turned out.View attachment 223159

no it doesn't. it literally says autonomous driving trends. its talking about the industry not mobileye roadmap.

36 mins

here is mobileye roadmap.

 

[bonnie]

no it doesn't. it literally says autonomous driving trends. its talking about the industry not mobileye roadmap.

here is mobileye roadmap.

Curious. You're in MIchigan ... what do you do for a living?

 

[Bladerskb]

Curious. You're in MIchigan ... what do you do for a living?

i'm a software engineer. what about you?

 

[bonnie]

i'm a software engineer. what about you?

I was early in my career. Aviation to start with. Moved into medical device, etc.

But of course, that isn't the point here . Are you in Detroit? Usually when someone has such polarized views, there is a reason. You don't like Tesla, it's not because of a negative ownership experience, and you seem to have a lot of info on other companies. So obviously I'm curious if you're working for one.

 

[bhzmark]

Even if the manufacturers are not enabling AP features, these cars are still learning and collecting data. It is approximately 15 million cars by 2016 year end. This already includes 47 vehicle models from 11 OEMs, to be produced in 2018-2026.

Name the make and model of cars using Mobileye now for sale.

There are dozens of car manufacturers with level 2 driver assistance using mobileye.

Name the make and model of cars using Mobileye now for sale.

 

[Bladerskb]

I was early in my career. Aviation to start with. Moved into medical device, etc.

But of course, that isn't the point here . Are you in Detroit? Usually when someone has such polarized views, there is a reason. You don't like Tesla, it's not because of a negative ownership experience, and you seem to have a lot of info on other companies. So obviously I'm curious if you're working for one.

I do like tesla actually and had a model 3 reservation and cancelled once I researched and discovered that elon claim that he will produce 200k models 3 by the end of 2017 was all fluff and I wont be getting by car until the end of 2018.

believe it or not but I bought into the tesla hype until I started doing my own research and started working with autonomous self driving systems. then I discovered that a lot of things elon says is pure hype. no I do not currently work in the automotive field, but that could change in the coming months.

 

[Bladerskb]

Name the make and model of cars using Mobileye now for sale.



Name the make and model of cars using Mobileye now for sale.

About Mobileye | Mobileye

over 10 million vehicles worldwide.

in Q1 of 2015 for example, its eyeq3 chip was installed in 816,000 new cars

Mobileye wins 2 contracts for its EyeQ4 chip

 

[GHammer]

no it doesn't. it literally says autonomous driving trends. its talking about the industry not mobileye roadmap.

36 mins

here is mobileye roadmap.

Right, So the CTO of mobileye gives a presentation on "industry trends" but doesn't use his own company as the example? In other presentations where he used that slide he specifically referenced the Q3.

 

[Bladerskb]

Right, So the CTO of mobileye gives a presentation on "industry trends" but doesn't use his own company as the example? In other presentations where he used that slide he specifically referenced the Q3.

one he's not giving a presentation on industry trends. two that is just one slide, 30 seconds out of his 30 minutes talk.
if you actually watched his entire talks and not grabbed and posted an image you got off of a google search then you would be more knowledgable.

 

[GHammer]

one he's not giving a presentation on industry trends. two that is just one slide, 30 seconds out of his 30 minutes talk.
if you actually watched his entire talks and not grabbed and posted an image you got off of a google search then you would be more knowledgable.

I have and I didn't get that from a google search it was taken from other one of Prof. Shashua's talks that I have seen in it's entirety. You are contradicting yourself, in one post you say he his taking about the "industry" not mobileye and the next you say he isn't. Why would he put up a specific year roadmap and not be talking about Mobileye.