How will Tesla demo FSD?

[S4WRXTTCS]

Is Super Cruise considered a publicly available version of Cruise? Or are they considered entirely different? If they're considered part of the same thing, then I'm fairly unimpressed with Super Cruise. Geo-fenced lane keep assist is neat, but it doesn't exactly knock my socks off. But, based on autonomy filings in CA, I do know that people love crashing into Cruise's vehicles while they're stopped at intersections.

Super cruise, and Cruise are wildly different things.

Entirely different teams, and entirely different purposes.

Super cruise was being worked on well before GM even bought Cruise.

The one interesting thing about Cruise was they originally were working on a kit that was supposed to give normal cars a highway automation add-on kit. But, they went away from that idea. Probably discovered it was a lot more lucrative to go full on, and simply get bought out for hundreds of millions.

Hence the biggest problem with self-driving cars is there is just too much money in it. Waymo was paying employees so much money that they could simply say screw you, and walk away after a couple years to start their own thing.

Even Uber of all companies managed to swindle a billion dollars recently for their self-driving car efforts.

On April 22nd the money is going to be rolling into Tesla.

I'm an excited by what Tesla is going to show. I can't wait to see the test drives, and even the silly little pointless demo.

I can't wait.

But, I still know it's a dog and pony show. In business you have to put on these events to show things well before they're ready.

You gotta excite the money people. It' just a necessary evil of life.

 

[electronblue]

You gotta excite the money people. It' just a necessary evil of life.

And there we have, very likely, the reason for 22nd April 2019, just as it was for October 19th 2016.

It may not have that much to do with the actual progress as with the need to excite the money people @diplomat33.

I’m sure it will be a good show. October 19th 2016 was a good show too.

Here is one reaction from October 19th:

 

[diplomat33]

It may not have that much to do with the actual progress as with the need to excite the money people @diplomat33.

Yes, exciting the money people is a big reason for doing the event. I don't deny that. Heck it is in the name of the event, "Autonomy Investor Day". But I don't think it is the ONLY reason. Exciting Tesla customers who have waiting very patiently for FSD is another reason for it. Exciting the general public so that they get interested in Tesla is another reason. But I also think that Tesla feels that they have reached a milestone in their FSD work (AP3 and "feature complete") that they now have enough material to do this kind of event. Like an author who has written 2-3 chapters of a book so they feel that they have enough to show a publisher to get an allowance to finish the book.

 

[Bladerskb]

Furthermore, I would argue that if Tesla is being intentionally secretive, why let Musk boast about FSD which only draws attention and scrutiny? Again, if Tesla is being intentionally secretive in order to hide their lack of progress, it would be better to keep a low profile and talk about FSD. So yes, Tesla is secretive in their FSD work but not intentional IMO.

Except we have proof that Elon lives off of hype. Everything he does he hypes to the uttermost and will use ambiguity and lies to do it, not the actual truth. Think about it, everything about FSD and the fact they are 3 years ahead of the competition and "game, set, match" is literally a lie. Nothing is based on actual truth. Why is that? Why can't they base anything on actual truth?

 

[boonedocks]

Yes, exciting the money people is a big reason for doing the event. I don't deny that. Heck it is in the name of the event, "Autonomy Investor Day". But I don't think it is the ONLY reason. Exciting Tesla customers who have waiting very patiently for FSD is another reason for it. Exciting the general public so that they get interested in Tesla is another reason. But I also think that Tesla feels that they have reached a milestone in their FSD work (AP3 and "feature complete") that they now have enough material to do this kind of event. Like an author who has written 2-3 chapters of a book so they feel that they have enough to show a publisher to get an allowance to finish the book.

I just want Tesla to demonstrate and show something significant to FSD so they can get on with upgrading all of us HW2.0 cars with some descent hardware.

 

[Bladerskb]

As consumers we don't really know where the failures within the ADAS are coming from. Are they failures of the vision system or some other part of the ADAS system?

Both you and Blader went out your way to make sure I was educated that MobileEye wasn't the complete ADAS vendor in most cases.

We do. Most people don't because most people don't research their questions but draw their own conclusions. But just a couple minutes of research will show that different automakers have different tier 1s which provide automakers with control algorithm to implement NCAP safety solutions (including AEB) and other ADAS features. This part isn't a blame game, its a fact.

Companies like Mobileye only provide the vision chip and software. The tier 1s have completely dropped the ball.

But, customers like myself don't really care.
We just want a system that has lots of capabilities, and we don't want a bunch of finger pointing.

True, which is why Mobileye is now trying to offer a complete end to end full stack system. But getting the automakers to buy in is the problem. They are lazy, lack ambition and are completely incompetent. They have settled for mediocrity for decades from Tier 1s.

Mobileye is now offering a full stack solution with surround cameras, their REM HD Map, their RSS system, and their own control algorithm (driving policy) which are both offspring of their Level 4 development. Which they call a Level 2+ system. It will handle highway driving and urban intersection driving (start/stop at traffic light/stop sign, etc) This would cost just a couple hundred dollars more.

But automakers are refusing, they just want plain mediocrity ADAS. They are the resistance to improving autonomy. So the blame lies squarely with them. They are the problem!

Within 6-12 months I think the any difference in the vision system will be irrelevant. My biggest concern is whether Tesla can make great strides on the driving logic front, and with maps.

Interesting enough Mobileye is already racking up millions of miles of HD Mapping and will be adding future Ford production cars to their REM HD mapping fleet. They are more in position to deliver what Tesla is trying to deliver. But alas none of the automakers want it.

My hope is that startups like Luicid Motors will deliver the full-stack Mobileye solution. Will buy day one.

I have a Model 3 right now for the reason you said (Tesla is more aggressive with rolling it out), and that I wanted a system where a single company had control over the entire ADAS system. So anything it didn't or didn't do would be completely the fault of that company. Assuming it's not some defect with the radar itself or cameras.

That's the exciting thing about Tesla. They're the only automotive maker that owns their own ADAS system. At least that I know of.

The frustrating part is that they don't parter up to get to the market with features faster.

GM is actually about to own their ADAS system full stack. Previously they were using mobileye vision system with their own interal development of control algorithm (not using any tier 1s).

Now they have brought everything in-house and are using cruise automation two forward facing camera system. You will see this in the 2020 Escalade that will launch early 2020 and will be available on the 2020 CT5 that was just revealed but the SC variant will be starting early 2020. Then there's another system called Ultra Cruise coming out probably in 2021 for urban driving.

 

[CarlK]

I don't think so. Super cruise is EyeQ3 plus super accurate maps. Cruise is a startup working on full autonomy that GM bought.
If Tesla is going to make a (possibly) fake demo they should try to top this one:

Yes Super Cruise and Cruise autonomous driving are two totally different things. The former is just your run of mill ADAS except GM did a couple pretty clever things, Lidar mapping and eye checking, both have nothing to do with the basic technology. Although GM did a great (meaning deceptive) marketing by tooting it as "hands free" and avoid to mention it's geofenced and will not work on most roads.

Yeah that video was so fakie I guess that's why they disabled the youtube comment section. They cleverly described the vedio as just to show what kind of situations an autonomous car would see to avoid any liabilities. You just have to give it to them.

I'd like to go on record and say that I believed MobilEye was a scam back in the 1999/2000 when they came out.

I only noticed that when I started to pay attention to the company when they were dealing with Tesla. A very dishonest company and frontman. They still were able to scam their way into having Intel to spend $15 billion plus whatever investment on them. You have give it to them too.

 

[Daniel in SD]

Yeah that video was so fakie I guess that's why they disabled the youtube comment section. They cleverly described the vedio as just to show what kind of situations an autonomous car would see to avoid any liabilities. You just have to give it to them.

How can you tell the video is fake? They claim in the description that it is real and 100% autonomous. Not saying it is or isn’t, just wondering how we will know if Tesla’s demo is fake this time when so many people were fooled last time.
Also, have you seen YouTube comments? It’s worse than TMC.

 

[MTOman]

We do. Most people don't because most people don't research their questions but draw their own conclusions. But just a couple minutes of research will show that different automakers have different tier 1s which provide automakers with control algorithm to implement NCAP safety solutions (including AEB) and other ADAS features. This part isn't a blame game, its a fact.

Companies like Mobileye only provide the vision chip and software. The tier 1s have completely dropped the ball.


True, which is why Mobileye is now trying to offer a complete end to end full stack system. But getting the automakers to buy in is the problem. They are lazy, lack ambition and are completely incompetent. They have settled for mediocrity for decades from Tier 1s.

Mobileye is now offering a full stack solution with surround cameras, their REM HD Map, their RSS system, and their own control algorithm (driving policy) which are both offspring of their Level 4 development. Which they call a Level 2+ system. It will handle highway driving and urban intersection driving (start/stop at traffic light/stop sign, etc) This would cost just a couple hundred dollars more.

But automakers are refusing, they just want plain mediocrity ADAS. They are the resistance to improving autonomy. So the blame lies squarely with them. They are the problem!

Interesting enough Mobileye is already racking up millions of miles of HD Mapping and will be adding future Ford production cars to their REM HD mapping fleet. They are more in position to deliver what Tesla is trying to deliver. But alas none of the automakers want it.

My hope is that startups like Luicid Motors will deliver the full-stack Mobileye solution. Will buy day one.



GM is actually about to own their ADAS system full stack. Previously they were using mobileye vision system with their own interal development of control algorithm (not using any tier 1s).

Now they have brought everything in-house and are using cruise automation two forward facing camera system. You will see this in the 2020 Escalade that will launch early 2020 and will be available on the 2020 CT5 that was just revealed but the SC variant will be starting early 2020. Then there's another system called Ultra Cruise coming out probably in 2021 for urban driving.

Just to make sure I understand what you're saying - no auto mfrs want to use Mobileye ADAS (because they want mediocrity) even though Mobileye is the best? And Mobileye ADAS is considered the best even though it's impossible to test drive a single car today that uses it because nobody has it? But IF they did it would blow Tesla Autopilot out of the water?

That logic seems air tight to me

 

[CarlK]

How can you tell the video is fake? They claim in the description that it is real and 100% autonomous. Not saying it is or isn’t, just wondering how we will know if Tesla’s demo is fake this time when so many people were fooled last time.
Also, have you seen YouTube comments? It’s worse than TMC.

It is edited. Only short clips when it worked are included. That's the problem with all "demo" videos. They are at best just for "demo" purposes but they are not real. Kind of like car chase scenes in movies. They are not real even real drivers were driving real cars.

 

[Electroman]

So basically you're saying that all FSD demos are a lie.

I will say it here for the 18th time, if you had paid any attention.

I don't care about any demo videos from anyone including Tesla. Hollywood makes better videos.

I want CUSTOMER TESTIMONIALS. In the dozens and hundreds. I have that in plenty for Tesla Autopilot 2.5 - over 386 Youtube videos by John Tesla-owner Doe driven by AP in highways, back roads, windy mountains, rain etc.. . SHOW ME customer testimonials for Mobile EyeQ.

Here is mine: My Model 3 AP is so bloody damn good, it drives me close to 90% of my travel. I really don't care for the rest 10% to make it to FSD. I am very satisfied with AP in its current form.

 

[croman]

No one should be happy with AP in its current form. Its only OK. Even Elon, the HypeMaster Supreme, understands they need to rapidly deliver immediate progress.

 

[Daniel in SD]

I want CUSTOMER TESTIMONIALS. In the dozens and hundreds. I have that in plenty for Tesla Autopilot 2.5 - over 386 Youtube videos by John Tesla-owner Doe driven by AP in highways, back roads, windy mountains, rain etc.. . SHOW ME customer testimonials for Mobile EyeQ

EyeQ is used in GM Super Cruise so I'm sure you can find some testimonials on that. Apparently it works pretty well but it is limited in which roads it allows you to enable it for.
Waymo is supposedly lifting its NDA for customers of the Waymo One service so it should be very interesting to hear how well it actually works.

Here is mine: My Model 3 AP is so bloody damn good, it drives me close to 90% of my travel. I really don't care for the rest 10% to make it to FSD. I am very satisfied with AP in its current form.

My testimonial is that it's the best cruise control I've ever driven. Unfortunately it seems very far from any sort of full autonomy. I haven't seen any testimonials about FSD probably because only employees have used it and it's under NDA.

 

[Malbers]

It is edited. Only short clips when it worked are included. That's the problem with all "demo" videos. They are at best just for "demo" purposes but they are not real. Kind of like car chase scenes in movies. They are not real even real drivers were driving real cars.

Cruz is driving around San Francisco every single day and it's not a one-off highly produced event. When will it be ready? I don't know but I doubt they'll sell the full package to the general public anytime soon; however, Uber and Waymo should be concerned.

Personally I think remote operators will be required for some time but we'll see

 

[diplomat33]

Except we have proof that Elon lives off of hype. Everything he does he hypes to the uttermost and will use ambiguity and lies to do it, not the actual truth. Think about it, everything about FSD and the fact they are 3 years ahead of the competition and "game, set, match" is literally a lie. Nothing is based on actual truth. Why is that? Why can't they base anything on actual truth?

Let's separate fact from hype. Tesla put out this official statement: "Tesla is making significant progress in the development of its autonomous driving software and hardware, including our FSD computer, which is currently in production and which will enable full-self driving via future over-the-air software updates,"

So according to Tesla, they have made "significant progress" on FSD, enough progress in fact to warrant a big media event on Monday. I have no reason to doubt that statement. I firmly believe that we will see a lot of FSD progress by Tesla on Monday. The talk of "3 years ahead" and "game set and match" is just how Elon and the media too are characterizing the "significant progress". Don't underestimate the role of the media either. They are creating a lot of hype too. It's not just Elon. Yes, the characterizations might turn out to be hype. In fact, there is a real danger that when you set the bar so high, that the "significant progress" won't live up to the hype. But as a Tesla owner, I am much more interesting in Tesla making progress and seeing what that progress is, than what Elon says. At the end of the day, Elon claiming that they are X months away from probably doing L4 or whatever is just talk. What matters is the software update I get in my car and what it can do.

 

[Bladerskb]

Just to make sure I understand what you're saying - no auto mfrs want to use Mobileye ADAS (because they want mediocrity) even though Mobileye is the best? And Mobileye ADAS is considered the best even though it's impossible to test drive a single car today that uses it because nobody has it? But IF they did it would blow Tesla Autopilot out of the water?

That logic seems air tight to me

I'm not sure what you are getting at here. They do use mobileye products, but settle for the cheapest barebone features from both mobileye and tier 1s leveraging mobileye chips. Rather than opting for the high-end features. All they want to do is be compliant with NCAP so they can keep selling their cars. Aka look at FCA paying Tesla millions to avoid EV penalties. If tomorrow a law was passed requiring every car to have Level 2+ ADAS and meet a certain performance standard then they will actually act.

Amnon has said it himself, unless you enact regulation to push the automakers they won't move.

Mobileye themselves have lobbied for regulations, recently the Intelligent Speed Assistance law, or the new NCAP for wide view that goes into affect in a-couple years. Mobileye heavily lobbied for those. Why? the only way you will get advanced automation from trad automaker is by forcing their hand through regulation.

New Cars in Europe Will Be Prevented From Speeding

The legislation will require all new vehicles put on the market to be equipped with Intelligent Speed Assistance (ISA) technology. The devices can use sign-recognition video cameras or GPS-linked data to automatically limit the speed of a vehicle.

 

[electronblue]

Just to make sure I understand what you're saying - no auto mfrs want to use Mobileye ADAS (because they want mediocrity) even though Mobileye is the best? And Mobileye ADAS is considered the best even though it's impossible to test drive a single car today that uses it because nobody has it? But IF they did it would blow Tesla Autopilot out of the water?

That logic seems air tight to me

Your ignorance unfortunately betrays you. MobilEye sells a vision chip, which is then integrated by the car makers as they see fit. The car makers often buy this integration work as solutions from so called tier 1 suppliers — outsourcing it if you will. It is the tier 1 suppliers and the traditional car makers that lack ambition and that shows in the lackluster integration they’ve so far attempted. MobilEye’s vision chips offer great range of reliable data from the world, far surpassing what Autopilot 2 feeds AP2/2.5 hardware today, and these boring companies use a miniscule amount of it.

I don’t necessarily share @Bladerskb ’s pessimism about MobilEye’s latest full stack offering getting used in the industry — it is new and the industry moves slowly. It may well still find its uses. What matters more in the current lackluster analysis is that for the past decade+ when four generation of EyeQ chips have been used by the industry, MobilEye has sold a vision solution only — the integration work has had to come from others, who have not been ambitious...

The minute someone bold started making use of MobilEye technology... guess what was born...?

Yes indeed, Autopilot 1.

That said, signs are car responsible driving in the form of Level 3 solutions using MobilEye hardware are coming from a few manufacturers, so that is not nothing. But so much more could happen were they adventurous enough as Tesla is, because MobilEye’s vision is far more all-seeing and robust today.

It is probably MobilEye’s Level 4 taxi fleet project that will first show us their true capability.

 

[CarlK]

Cruz is driving around San Francisco every single day and it's not a one-off highly produced event. When will it be ready? I don't know but I doubt they'll sell the full package to the general public anytime soon; however, Uber and Waymo should be concerned.

Personally I think remote operators will be required for some time but we'll see

Maybe so but Google/Waymo had test cars running around Mountain View everyday years ago and still does not have real things to show. You have no idea how these were run unless there are third party observers who did not sign NDA in the car.

Remote operators probably will always be required for truly "driverless" cars. At least someone needs to intervene in case the car is stuck because of accident, weather condition, etc..

 

[MTOman]

Your ignorance unfortunately betrays you.

Which is precisely why my post was full of questions, I'm ignorant about Mobileye. To you that's unfortunate, to me that's exactly how one should respond to being ignorant about something

MobilEye sells a vision chip,

Mobileye is now offering a full stack solution with surround cameras, their REM HD Map, their RSS system, and their own control algorithm (driving policy)

Excuse me for being ignorant once again, but isn't what @Bladerskb is describing an ADAS, or autopilot-like, system?

They do use mobileye products, but settle for the cheapest barebone features from both mobileye and tier 1s leveraging mobileye chips.

Because all they want is mediocrity, right? Seriously, what's the real reason they're developing their own systems if Mobileye is so much better?

 

[electronblue]

Which is precisely why my post was full of questions, I'm ignorant about Mobileye. To you that's unfortunate, to me that's exactly how one should respond to being ignorant about something

Excuse me for being ignorant once again, but isn't what @Bladerskb is describing an ADAS, or autopilot-like, system?

Yes, so basically it is like this:

- Up until recently, MobilEye only sold vision chips for ADAS features (integrating these was up to the manufacturer), indeed Tesla made Autopilot 1 like this (this has gone on since the noughties)
- Now MobilEye also sells a full ADAS solution that is not yet deployed in any consumer car, this is quite recent and based on the EyeQ4 chip that launched last year that features both vision + driving policy option
- MobilEye also develops a full self-driving system (Level 4) in-house, this is so far not sold to anyone but is based on the same chips of course

Because all they want is mediocrity, right? Seriously, what's the real reason they're developing their own systems if Mobileye is so much better?

Car makers have various reasons of course but one main reason is their tight integration with their supplier network. Traditionally car manufacturers (in recent decades) have focused on design and a small subset of the car themselves (like the engine) and bought the rest from their supplier network. So far ADAS has been more about checking a box on the feature list than something to worry about so they’ve been shopping what their supplier network can easily supply them. This is why the ADAS systems have been quite lackluster...

The another reason is that most car manufacturers do have in-house programs for full autonomy (using MobilEye chips) and car responsible driving at SAE Levels 3 and above. They are putting more resources in these but the plans to deploy those to the consumer fleet are seriously lagging. In the meanwhile they are letting their supplier network do their mediocre dance in the ADAS space.

Mind you, this mediocrity does not always mean ”worse than Tesla” or anything of the sort. Many of these ADAS systems do have features Tesla owners can currently dream about. But the significant issue is that there is a lack of boldness when it comes to actual self-driving in this setup. Tesla has this boldness and that is one reason Autopilot 1 both succeeded and, well, also failed a little at the same time. But it does set Tesla apart and Tesla made great use of MobilEye’s 2014 technology...