Autonomous Car Progress

[powertoold]

You are still missing my real point: we should judge ME based on their cutting edge, not based on what legacy automakers have deployed.

My opinion is we should place much more weight on what's been deployed, as we've learned from Tesla's "cutting edge" fsd videos. The best real-world evidence we have is from systems that are used by laypeople / non-affiliated users.

 

[diplomat33]

My opinion is we should place much more weight on what's been deployed, as we've learned from Tesla's "cutting edge" fsd videos. The only real world evidence we have is from systems that can be used by laypeople.

Ok. So make sure to check out Super Vision when it is deployed. I expect it will be way better than the current L2 on a lot of legacy automaker cars.

 

[powertoold]

Ok. So make sure to check out Super Vision when it is deployed. I expect it will be way better than the current L2 on a lot of legacy automaker cars.

For sure, we love videos of good fsd or adas performance. China is fairly restrictive about videos in general, so hopefully we’ll get some actual user videos.

 

[Bladerskb]

Having worked on large software products from ground up, there's a huge difference between something that you can demo vs something you can deploy to laypeople / users. We always have to take that into account when dealing with fsd software.

I'm sorry but tesla fans have paraded around and championed "Tesla FSD" for years. Now you are all of a sudden claiming there is a difference between a demo product and a production system. There's 77 or whatever number of people with literally a demo product which you are championing. Yes FSD Beta in the hands of those 77 or so people is a demo product, not a production system. That is smaller and worse than the test team at other SDC companies.

Other companies like Mobileye and Huawei are about to release a REAL production door to door system in a matter in months. While your Elon will continue to promising exponential quantum leap in two weeks.

Secondly you admitted you have zero experience in ML which is not surprising from your statements.
No offense but you keep saying you worked on "software products" or "software development" which could be an umbrella for anything from a QA guy to a Data Analyst.

 

[diplomat33]

What I find most interesting about Waymo's challenges is that despite having everything fully mapped, they're still avoiding many unprotected lefts and lane merges, to name a few. It really doesn't bode well for the other fsd developers with similar approaches hoping to catch up to Waymo.

I don't think mapping is the issue. Waymo has no problem mapping intersections. And Waymo does not avoid unprotected left turns when there is a safety driver, they only avoid some situations when it is a driverless ride. I think it is more that traffic or conditions could make it a bit riskier and Waymo does not want to take the chance with a driverless ride.

After all, when you are testing with a safety driver, you can take some risks. But when it is a driverless ride with a mom in the back seat with her kid, going to soccer practice, it makes more sense to take the safest route, especially since a robotaxi can choose any route it wants as long as it gets there in a reasonable amount of time. We've even discussed the fact that Tesla might do the same thing if they ever deployed a L4 ride-hailing service. We know there are unprotected left turns that FSD Beta won't do but Tesla could still deploy L4 if they rerouted to avoid them.

I would also add that the 5th Gen on the i-Pace is light-years better than the 4th Gen on the Pacificas that we see in Chandler. I believe when Waymo does deploy the 5th Gen FSD, we will see a quantum leap in improvement.

 

[powertoold]

I would also add that the 5th Gen on the i-Pace is light-years better than the 4th Gen on the Pacificas that we see in Chandler. I believe when Waymo does deploy the 5th Gen FSD, we will see a quantum leap in improvement.

Will believe when we see it deployed!

 

[stopcrazypp]

Other companies like Mobileye and Huawei are about to release a REAL production door to door system in a matter in months. While your Elon will continue to promising exponential quantum leap in two weeks.

Let's not count the eggs before they hatch, especially when talking about volume releases. I'm getting deja vu on the Audi's promise of release of L3 in 2017/2018. 3-4 years later the only L3 we see is still only a 100 vehicle lease only fleet from Honda.

As for Mobileye, there's already a discussion above about the lead times from demo to volume consumer release.

For Huawei, my confidence in the claims of that company went down like a rock given their deception on HarmonyOS/HongmengOS (this article in particular being very revealing):
Huawei’s HarmonyOS: “Fake it till you make it” meets OS development

 

[Bladerskb]

I still don't agree with what you're claiming about legacy auto implementations of Mobileye systems.

That's the whole point of Mobileye advertising EyeQ4 with a list of features. That means the chip comes with software features that automakers can rebadge or adjust to their own liking.

ZF and Mobileye Safety Technology Chosen by Toyota

www.intc.com

"Mobileye’s EyeQ4 applies enhanced computational capabilities with computer vision algorithms while rapidly processing information from the vehicle’s front-facing camera. With a variety of feature sets including vehicle detection from any angle and next-generation lane detection, EyeQ4 enables automakers to take a major step forward in autonomy, achieving the ability to support and enhance complicated driving tasks."

That's because the EyeQ comes with the vision NN but the driving policy is developed by the Tier 1s or automakers. That's why there's variety of difference in features and performance from various systems. You can use the chips to just do AEB if that's what you want.

But given there are no ADAS in consumer hands as advanced as Tesla's other than Enhanced Super Cruise (which as mentioned is only recently adding some features like auto lane change that Tesla had 6 years ago), it's pretty safe to say Mobileye haven't made dent yet (at least here in USA, in China there are a few examples of ADAS systems that are as good as Tesla's, but not necessarily using Mobileye, most of them are using Nvidia, from a quick search). No doubt their demos are impressive, but how long will it be before consumers can buy it and will it be affected by the customizations automakers may make that may make them inferior, like for example that Copilot one using EyeQ4?

Navigate on autopilot which was released in 2018 is what came with self lane change, so not 6 years ago for the self lane change. Auto lane change which was there with AP1 came to Super Cruise 2.0 last year. The latest update to SC adds the self lane change.

Nio pilot uses EyeQ4 and has Navigate on Nio Pilot, majority of Chinese automakers existing system use Mobileye. The only company that i know of that has successfully deployed a Nvidia based system is Xpeng.

Secondly you have Propilot 2.0 that is avaliable in Japan and in the US that uses EyeQ4.
Then you have Honda the L3 Honda legend which also has a navigate feature and most likely uses EyeQ4 but i can't confirm.
Finally you have none mobileye systems like Xpeng and the recent Toyota Teammate in Japan and available in the US this year which both have a navigate feature

 

[Bladerskb]

That's an interesting point. From what I see in that datasheet, the "SuperVision" is basically two EyeQ5 High chips and 11 cameras, but it doesn't explicitly say automakers can't customize parameters or limit how the vehicle control happens. Looking at the Copilot video, the failure to make the tighter turns can easily simply be from an automaker set limit on the amount of steering angle that can be used by the system. If this sort of customization is allowed in the SuperVision product, then the same kind of failures can occur even if they offer the entire stack (including the sensors) for perception, decisions, UI.

I see the same issue in videos of the Nvidia based systems that the Chinese automakers are using, they consistently fail miserably on tighter turns because it seems like it's hitting some sort of steering angle limit (and the software isn't able to work through the turn with repetitive smaller adjustments). It is not a failure of perception given the road seems to draw correctly and they have HD/MD maps of the roads (in far better detail than Tesla, as Tesla uses regular Baidu maps, which seems far inferior to the Amap maps that a lot of others are using).

The EyeQ vision chips doesn't do driving policy. They contain only perception NNs. Absolutely no planning or control algorithms. Ford's Copilot is either using driving policy developed by Ford or a Tier 1. Just like BlueCruise is using driving policy developed by Ford.

Mobileye for the first time is offering driving policy and its only happening through the supervision package.

 

[Bladerskb]

Let's not count the eggs before they hatch, especially when talking about volume releases. I'm getting deja vu on the Audi's promise of release of L3 in 2017/2018. 3-4 years later the only L3 we see is still only a 100 vehicle lease only fleet from Honda.

As for Mobileye, there's already a discussion above about the lead times from demo to volume consumer release.

For Huawei, my confidence in the claims of that company went down like a rock given their deception on HarmonyOS/HongmengOS (this article in particular being very revealing):
Huawei’s HarmonyOS: “Fake it till you make it” meets OS development

I was one of the first ones who called out Audi and BMW years ago when it was clear they were just in it for the PR. I even made a thread about it.
Audi Pilot Driving actually had a good system but management gobbled it up like they did the 4 other development project that came after Piloted Drive was shut down.
Most people think i'm biased, but i'm not. I have ripped into just about every traditional automaker out there.

Update on BMW's INEXT Level 3 Highway System in 2021 (not happening)

UPDATE: A few updates are on in order. Looks like Mobileye is no longer joint with BMW on driving policy although they have given BMW the blueprints to move forward. This change most likely came from the new Mercedes/BMW autonomous deal in FEB 2019. Not only that, as a result of this Amnon no...

teslamotorsclub.com

About the Huawei Phone OS, they were forced to make a phone in acouple months without Google Android so they forked the Android Open Source Platform (which alot of companies do) and rebranded it with plans to improve and differentiate as time goes. This is actually what makes them a tech company. An automaker wouldn't do that. They would take 5 years and then come out with absolutely nothing. Tesla did the same thing by ultilizing GoogleNet. Anyway this is different and meanless to the discussion. There is no good open source SDC software to fork.

Depends on what you define for "right". which is the crux of the argument (and why as I mentioned trying to pick "winners" in a comparison chart doesn't work given people have different criteria).

If "right" is releasing L5, then none of the players are "right" given none of them have reached L5 (or even just wide release L4). The basic paths being taken I see is:
1) Tesla's approach: work to release end-to-end L2, then improve reliability of that until it is to L4 then eventually L5 level. The idea is that they can convert their fleet at a flip of a switch as soon as the software is ready.
2) Others: release L4 in geofenced area relying heavily on HD maps, and achieve L5 by mapping the entire world or at least specific countries. Note that many companies don't have a explicit goal to achieve L5, they would be perfectly happy with L4 (robotaxies can operate just fine in geo fences), in which case the goal would instead be wide release L4 (releasing L4 in many more places than just experimental fleets, like Uber/Lyft's current coverage area vs when they originally started a decade ago).

I'm talking about Elon claiming level 5 in two years for the past 6 years and others either claiming Level 5 is impossible or that it will take a long long time, some mentioning 2030+.
But Elon kept claiming he will do it in 2018.

Propilot is probably the biggest disappointment. I remember all the hype about it being L4 (from your post in 2017 "Complete autonomous driving for all driving situation on the highway.") and then the final release isn't even L3 and it hasn't even reached US shores yet as of now. From what I can find, it was released in a limited manner in Japan in late 2019 for some Skylines. That's the biggest problem with all of Mobileye's solutions, it takes forever to reach the consumer.
Four Upcoming Self Driving Level 3 Cars by 2019

That system is based on EyeQ4, and from a quick google, Mobileye had engineering samples already back in 2015 (they may have demoed an early release earlier than that even, given you follow Mobileye closely, perhaps you know):
Moving Closer to Automated Driving, Mobileye Unveils EyeQ4® System-on-Chip with its First Design Win for 2018

And here 6 years later we are still waiting for it to hit US shores in volume. Does that mean we won't see any significant release of the things demoed by Mobileye recently until 2027? A lot can change by then.

Neither nissan nor me ever promoted the propilot 2.0 as L4, but always as L3 in every related marketing. Even then the system as it is was never released. I'm not even talking about whether its L2 or L3. I mean the original system had 4 lidars, 12 surround cameras. The release hardware had 0 lidars and 3 forward cameras.

It definitely was disappointing. But its par for course for traditional automakers. I made this post over 2 years ago and its still accurate. There's a reason there's still no automaker with a reliable OTA update system after years on years. Yet startups perfect it from day one.

Note that mercedes actually ended up giving up and going completely Nvidia for both hardware, software, ADAS and AV.
VW is still stuck creating teams, fumbling the development, nuking everything and creating another team. Rinse and repeat.

Update on BMW's INEXT Level 3 Highway System in 2021 (not happening)

UPDATE: A few updates are on in order. Looks like Mobileye is no longer joint with BMW on driving policy although they have given BMW the blueprints to move forward. This change most likely came from the new Mercedes/BMW autonomous deal in FEB 2019. Not only that, as a result of this Amnon no...

teslamotorsclub.com

Its not even just because of safety risk. they absolutely have no ambition. You can hand them a L5 self driving system today free and they will find a way to fumble the roll out and spend 3 years just deciding what to do with it. Then spend another 3 three rolling out 100 cars in a city. safe to say they are hopeless.​

​

Its not the engineers its the 80 years old business suit heads calling the shots.​

​

The only hope are start-ups. These two articles are good reads...​

​

Has Mercedes just given up?​

​

The Frightening Low Ambition Level of German Carmakers in Regards to Autonomous Driving​

The reason for the long deployment time is due to the trad automakers. Mobileye already came out and said it takes traditional automakers and tier 1s 3-4 to integrate. EyeQ4 went to production in Q4 2017 and it took NIO till the first half of 2018 to integrate and deploy it. So less than a year. Yet it takes someone like Ford and GM 4 & 3 years respectively to deploy EyeQ4 and Ford with crap driving policy. For EyeQ5, Geely's new Zeekr brand is integrating and deploying less than 1 year after its production date.

So blame the traditional automakers. However the eventually released propilot 2.0 in Japan is coming to the US this year.
Anyway, If you want the latest and greatest Mobileye system, you will only find it on EV startup cars not traditional automakers.
I mean you will literally see the latest and greatest supervision on the Zeekr this year.

 

[JHCCAZ]

...Yes FSD Beta in the hands of those 77 or so people is a demo product, not a production system. That is smaller and worse than the test team at other SDC companies.

But we're not seeing unofficial user-experience test videos, as the system develops warts and all, from those SDC companies. As I've said before, this is quite a remarkable and brave thing that's happening before our eyes, (and I don't much depreciate that just because Tesla asks for no Live-Stream, that's a reasonable limitation for several reasons and apparently theyre not vetting or censoring the content before public upload after each drive). AFAIK this is unheard of from any other in-development (beta) SD - except I guess comma which is kind of in perpetual development and open-source, and strangely gets very little share of attention here.

...Secondly you admitted you have zero experience in ML which is not surprising from your statements.
No offense but...

I noticed this recent turn of the usual heated exchange. This is a forum open to thoughtful debate, information and commentary. Most but not all participants have a somewhat technical background. However it is not a requirement, nor should it be a litmus test for validity of a presented argument. It is a bad enough though common fallacy to dismiss opposing arguments based on appeal to 3rd-party authority. Authority or experience in themselves prove nothing, other than the argument might be worth an investment of one's limited time when deciding which ones to examine. Worse is to extend it to ridicule of the 2nd-party, and IMO worst of all is to appeal to one's own 1st-party authority. We should all be evaluating arguments based primarily on intrinsic merit and agreed facts (though itself a further basis of debate).

Self-driving is a challenging, fascinating and unsolved problem. We're all entitled to our views on the most promising approaches and players. But after it happens, I wouldn't believe anyone, even the person who got it right in the end, who claimed to have understood the solution all along. It is thus not a logical nor credible approach to base one's debating points on the Machine Learning, programming or automotive-engineering qualifications of the debating parties.

I'll leave you with one of the better quotes on this, from John Locke who, it's safe to say, was unbiased regarding self-driving competitors. And no I don't think this is a contradictory appeal to authority by quoting the words of a great man. His words are not important because he was famous; he is famous because his words are important.

(I)t is not what the man of science believes that distinguishes him, but how and why he believes it. His beliefs are tentative, not dogmatic; they are based on evidence, not on authority or intuition.​

 

[powertoold]

Most but not all participants have a somewhat technical background. However it is not a requirement, nor should it be a litmus test for validity of a presented argument. It is a bad enough though common fallacy to dismiss opposing arguments based on appeal to 3rd-party authority. Authority or experience in themselves prove nothing, other than the argument might be worth an investment of one's limited time when deciding which ones to examine. Worse is to extend it to ridicule of the 2nd-party, and IMO worst of all is to appeal to one's own 1st-party authority. We should all be evaluating arguments based primarily on intrinsic merit and agreed facts (though itself a further basis of debate).

I don't know how much interest you have in AVs, but this Mobileye interview is the basis of my negative judgment of their approach. The video begins at a bunch of illogical statements by Mobileye's leader, Amnon:

 

[SMAlset]

Not sure where to post so move if deemed better placed:

Husband saw this Nvidia Test and Development Ford vehicle while out today. Thought you guys might like a peek. Interestingly it seems to record audio too from the display sign. I thought a lot of states had restrictions on audio recordings. Maybe they just want to pick up passerby comments on the car?

The car is a Ford Fusion Platinum. Assume their BlueCruise driver assist system being tested on this model. Saw a The Drive article from April mention BlueCruise would be available on the F-150s Platinum line among others so the Platinum designation stood out on this model.

 

[hgmichna]

I'd also like to see some kind of visual and audible signal given when a vehicle is moving without a driver

How about a red flag on a long pole? That's been demanded before (for cars without a horse).

 

[Matias]

That is not Mobileye's system. That is Ford's system, powered by the Mobileye chip. That is an important distinction. That is the mistake that you and @powertoold are making. These are NOT Mobileye's in-house system. They do not represent what Mobileye has. They represent what the legacy automakers have in terms of ADAS.

Mobileye has progressed a lot. Mobileye's system is called Super Vision. It is what you see in the NYC and Munich FSD drives.

It is true different automakers use same MobilEye’s hardware completely different. Tesla’s AP1 has Mobileye’s EyeQ3 and some other carmakers also have it. Yet Tesla used it better than others.

 

[Dan D.]

...
I'd also like to see some kind of visual and audible signal given when a vehicle is moving without a driver
...

How about a red flag on a long pole? That's been demanded before (for cars without a horse).

Have you seen Smart Summon work in a busy parking lot? There's lots of videos. A good number of them end with the car jammed in the way blocking other cars. The owner has to run over, get in embarrassed and drive away.

Yes, for now I think you're right. It would be a good idea to preceed your car by walking with a red flag or just drive it 'autonomously' on Smart Summon from inside the car. If it's going to drive terribly, it is a hazard.

 

[Bladerskb]

Not sure where to post so move if deemed better placed:

Husband saw this Nvidia Test and Development Ford vehicle while out today. Thought you guys might like a peek. Interestingly it seems to record audio too from the display sign. I thought a lot of states had restrictions on audio recordings. Maybe they just want to pick up passerby comments on the car?

The car is a Ford Fusion Platinum. Assume their BlueCruise driver assist system being tested on this model. Saw a The Drive article from April mention BlueCruise would be available on the F-150s Platinum line among others so the Platinum designation stood out on this model.

nah lots of people use variety of cars for their testing platform. This isn't bluecruise.

 

[Matias]

Have you seen Smart Summon work in a busy parking lot? There's lots of videos. A good number of them end with the car jammed in the way blocking other cars. The owner has to run over, get in embarrassed and drive away.

Yes, for now I think you're right. It would be a good idea to preceed your car by walking with a red flag or just drive it 'autonomously' on Smart Summon from inside the car. If it's going to drive terribly, it is a hazard.

It is surprising, that it is legal to let 2t projectile, which is controlled by alpha sofware, to move autonomously among pedestrians.

“Smart” summon has repeatedly crashed obstacles.

 

[powertoold]

It is true different automakers use same MobilEye’s hardware completely different. Tesla’s AP1 has Mobileye’s EyeQ3 and some other carmakers also have it. Yet Tesla used it better than others.

Mobileye provides a base feature set. I guess the carmaker can improve on it if they wanted? But then again Tesla was trying to improve it to the point where they take control of it, and that's when they split.

 

[diplomat33]

But then again Tesla was trying to improve it to the point where they take control of it, and that's when they split.

The split did not happen because Tesla was trying to improve ME's system. ME broke up with Tesla because they felt that Tesla was misusing their system to do FSD that it was not designed to do and compromising safety:

On Wednesday, Mobileye revealed that it ended its relationship with Tesla because "it was pushing the envelope in terms of safety." Mobileye's CTO and co-founder Amnon Shashua told Reuters that the electric vehicle maker was using his company's machine vision sensor system in applications for which it had not been designed.

"No matter how you spin it, (Autopilot) is not designed for that. It is a driver assistance system and not a driverless system," Shashua said.

Source: Mobileye spills the beans: Tesla was dropped because of safety concerns