Autonomous Car Progress

[Bladerskb]

More videos of Huawei's 2021 Door to Door Autopilot system rides.

 

[Bladerskb]

Zeeker 001 with Mobileye's SuperVision also with Door to Door Urban autonomous piloting

"In 2022, we will be the first in the industry to launch the urban autonomous driving function, so that autonomous driving can be started from your doorstep." - Zeeker

Sensors

• 7x 8-megapixel long-distance high-definition cameras​
• 1x 250 ultra-long sensing millimeter wave radar​
• 12x short-range ultrasonic​
• 4x short-range surround view HD cameras​
• 2x in-car monitoring cameras​
• 1x outside monitoring camera​
• 1x rear streaming media camera​

Full scene perception
Full scene prediction
HD REM Map
RSS (responsibility-sensitive safety models.)
Skeletal recognition (algorithm that can recognize pedestrian movements. hand stop, waving go, on phone, etc)

NZP high-speed autonomous navigation: On high-speed and express roads covered by high-precision maps, vehicles will automatically complete functions such as on-ramp, active cruise, active overtaking, and early lane change based on the navigation path.

NZP City Autonomous Pilot: It can identify traffic light information and realize automatic assisted driving in an urban open road environment.

AVP automatic parking In unmanned scene: remotely operated by mobile phone, the vehicle can automatically cruise to the designated parking space and park in and out of the parking space to cruise to the designated pick-up point.

 

[stopcrazypp]

Joshua Brown exaggerated the capabilities of Autopilot. He was not well aware of limitations. Contrary to popular belief white truck against bright sky had nothing to do with the problem. Mobileye only trained for back and front of vehicles, so it was not designed or trained for cross traffic. Had Joshua Brown known this, he might be alive today.

My point is he was well aware the car could disengage at anytime, but decided to watch a video on a portable DVD player anyways. Knowing the specific about not designed for cross traffic is irrelevant, as he was not looking at the road anyways.

 

[hgmichna]

In safety teaching, ie. the "Swiss cheese model of safety", Tesla's approach with the UI with L2 systems is a big hole in the cheese. It allows people's errors and misuse to result in a consequence. The reason is because the user interface is designed to be a Level 4 UI, not a L2. "Pling plong, Now the car has control, you have not".

That is far from the truth, because in Autopilot mode the driver can always instantly brake, steer, and accelerate. The human driver is always in full control.

They need to change the UI, we can't continue to sacrifice people to make autonomous progress.

Whom have we sacrificed? Has anybody been killed by the autopilot? Or have people killed themselves by misusing the autopilot? These are two very different things.

 

[FloridaJohn]

An article today over at Ars Technica about the state of autonomous driving.

Who will win the self-driving race? Here are eight possibilities

The article is written in a way to give odds to the eight possibilities. Here is the summary:

1. Waymo wins (20 percent)
2. Another robotaxi company wins (25 percent)
3. Tesla (and Comma.ai) wins (5 percent)
4. Mobileye (and its partners) wins (10 percent)
5. China wins (15 percent)
6. Self-driving trucks win (5 percent)
7. Delivery robots win (10 percent)
8. Somebody else wins (10 percent)

Of course interest on this forum will primarily be Tesla, so here is what they say about them:

This will make Tesla fans mad, but I think it's true: Tesla is a long shot.

The bullish case for Tesla is that it has access to a vast trove of real-world driving data harvested from customers' vehicles. If you think limited training data is a major bottleneck for improving self-driving algorithms, then this might be a significant advantage. Tesla CEO Elon Musk also has a bigger appetite for risk than most of the other companies working on self-driving technology. Musk's willingness to put unproven technology on public roads may accelerate Tesla's progress even as it creates a greater risk of fatal accidents.

On the other hand, Tesla has significant disadvantages. The company's business model—selling cars to end users—puts lidar sensors and high-density maps financially out of reach. Elon Musk has tried to spin this as a positive, calling lidar a "crutch." But the fact remains that almost every other company is using lidar and HD maps because it believes they are helpful.

More fundamentally, it's hard to watch videos of Tesla's software in action and conclude that Tesla is in a leading position—or even that it is catching up to the leaders. Tesla's unfortunately named "full self-driving beta" software routinely flubs scenarios that Waymo's cars have been able to handle for years.

And here is the conclusion of the article:

Another possibility is an "everyone wins" scenario. Maybe self-driving isn't one problem but a bunch of individual problems, each requiring its own carefully designed technology stack. Maybe driverless taxis, driverless vehicles, and driverless trucks are three distinctive markets that will each be dominated by its own group of companies. And maybe early versions of the technology will require so much human supervision that we'll have a slow and gradual transition from human-driven vehicles to driverless ones. I don't think this is a likely outcome, but it's possible.

Another possibility is that nobody wins: maybe self-driving is an even harder problem than people appreciate, even after years of setbacks, and it will take decades, rather than years, to get working. In that case, we could have yet another "AI winter" where a lot of companies scale back their research in this area. Again, I don't think this is likely, but it's a possibility.

 

[DanCar]

Waymo doesn't have a chance since they are risk adverse from hurting people. Although I would much rather be wrong and see Waymo widely succeed. Waymo does have a chance of success with long haul trucks on freeways. Waymo will succeed once leadership changes to someone less risk adverse like venture capatilists. Google / Alphabet is slowly selling off its stake in Waymo to fund operations.

Tesla is already winning charging $10K. Mobileye with its lengthy partner list will also win. And yes everyone wins when this technology matures in 10 years.

 

[diplomat33]

An article today over at Ars Technica about the state of autonomous driving.

Who will win the self-driving race? Here are eight possibilities

The article is written in a way to give odds to the eight possibilities. Here is the summary:

1. Waymo wins (20 percent)
2. Another robotaxi company wins (25 percent)
3. Tesla (and Comma.ai) wins (5 percent)
4. Mobileye (and its partners) wins (10 percent)
5. China wins (15 percent)
6. Self-driving trucks win (5 percent)
7. Delivery robots win (10 percent)
8. Somebody else wins (10 percent)

Of course interest on this forum will primarily be Tesla, so here is what they say about them:



And here is the conclusion of the article:

Good article. I mostly agree with their percentages. I agree that Waymo's slow expansion is the reason their odds are not higher. But I think 20% is a bit too low. Waymo's FSD is the best and with new leadership, I am more optimistic that they can expand faster. So I would put Waymo at 25%.

Mobileye has really good FSD, good FSD approach, lots of money and lots of connections to automakers to deploy their tech which they are in the process of doing. So I think 10% is a bit too low. I would put Mobileye at 15%, maybe even 20%.

I think the "everybody wins" options is very likely. If you look at most tech, there is rarely just one winner. Usually, there is a pioneer that starts the tech and then other companies that eventually dominate different markets. For example, Apple really invented the modern smart phone with the first iPhone and had a huge lead for awhile. And Apple is still a big force in the smart phone market but there are other phone makers like Samsung that have carved out a big piece of the market now. Likewise, Waymo has a big lead in L4 right now but I could see other companies like Cruise getting a big piece of the ride-hailing market, and Tesla or Mobileye getting their own piece of the consumer FSD market. I think the "nobody wins" is very unlikely because technology gets better and better. It is inevitable that we will solve FSD eventually.

But I always wonder what does "winning the FSD race" really mean. Does it mean being the first with widely deployed L4 or L5? And how many cars do you need to be considered "widely deployed"? Do the cars need to be consumer cars at a certain price range or can they be ride-hailing robotaxis? We have to agree on a definition of winning before we can really discuss who can win IMO.

 

[mikes_fsd]

I always wonder what does "winning the FSD race" really mean.

To me, it means who/which system saves the most lives on a daily basis!

In order to do that, you have to have it widely available/deployed, obviously!
Statistics says that Tesla Autopilot is already the winner, just no one wants to admit it!

 

[diplomat33]

To me, it means who/which system saves the most lives on a daily basis!

In order to do that, you have to have it widely available/deployed, obviously!
Statistics says that Tesla Autopilot is already the winner, just no one wants to admit it!

Except that saving lives does not require FSD (L4 or L5 autonomous driving). You can save a lot of lives with a L2 driver assist like Tesla is doing now. So that is a bad definition for "winning FSD" since it does not require FSD.

"Winning the race to FSD" must include achieving actual FSD (L4 or L5) in the definition.

 

[mikes_fsd]

which system saves the most lives on a daily basis!

Except that saving lives does not require FSD (L4 or L5 autonomous driving).

I clearly said "which system saves the MOST lives on a daily basis" and if you want to eliminate the most accidents you have to reach FSD!

Tesla, consciously and deliberately, chose to start saving lives ASAP by continually deploying new FSD features as ADAS! (and the stats keep proving the Tesla approach was/is right)

Where is the outrage that legacy OEM's are allowing humans to be slaughtered daily, only because they are afraid of a lawsuit!??

 

[diplomat33]

I clearly said "which system saves the MOST lives on a daily basis" and if you want to eliminate the most accidents you have to reach FSD!

Tesla, consciously and deliberately, chose to start saving lives ASAP by continually deploying new FSD features as ADAS! (and the stats keep proving the Tesla approach was/is right)

Where is the outrage that legacy OEM's are allowing humans to be slaughtered daily, only because they are afraid of a lawsuit!??

You don't need to reach FSD to save the most lives. FSD only saves lives if it is safer than human driving. FSD that is not safer than human driving would not save lives. So FSD itself does not automatically mean that you are saving the most lives. You can save the most lives with no FSD at all. You can save lives with just L2 ADAS if your ADAS is safer than human driving alone. Tesla is saving lives with just a ADAS that is safer than human driving alone, not with FSD.

 

[mikes_fsd]

You don't need to reach FSD to save the most lives. FSD only saves lives if it is safer than human driving.

At this point, I know you are being dense just for $#1t$ so, I will leave you with what I said in response to your 'I always wonder what does "winning the FSD race" really mean.' question:

To me, it means who/which system saves the most lives on a daily basis!

In order to do that, you have to have it widely available/deployed, obviously!
Statistics says that Tesla Autopilot is already the winner, just no one wants to admit it!

 

[Bladerskb]

At this point, I know you are being dense just for $#1t$ so, I will leave you with what I said in response to your 'I always wonder what does "winning the FSD race" really mean.' question:

To me, it means who/which system saves the most lives on a daily basis!

In order to do that, you have to have it widely available/deployed, obviously!
Statistics says that Tesla Autopilot is already the winner, just no one wants to admit it!

There are tens of millions of cars with L2 ADAS.
Super cruise for example has a safety record above AP (0 in 6 million miles). They might end up with incident/accident every 25 million miles as SC equipped cars expand. About 50x better than humans if we use Tesla's logic.

 

[mikes_fsd]

There are tens of millions of cars with L2 ADAS.

LMFAO, limited to some highways...

Reading comprehension on this board leaves much to be desired!

To me, it means who/which system saves the most lives on a daily basis!

In order to do that, you have to have it widely available/deployed, obviously!
Statistics says that Tesla Autopilot is already the winner, just no one wants to admit it!

 

[Bladerskb]

LMFAO, limited to some highways...

Reading comprehension on this board leaves much to be desired!

Actually most L2 ADAS available are NOT limited to highway.

BMW
Nissan
Volvo
Mercedes
etc..

Infact the only one limited to the highway is infact only GM's Supercruise.

 

[mikes_fsd]

Super cruise for example ....

Infact the only one limited to the highway is infact only GM's Supercruise.

Apparently you cannot even read what you wrote!

But, you're proving my point splendidly.

 

[Bladerskb]

Apparently you cannot even read what you wrote!

But, you're proving my point splendidly.

~40+ million L2 cars vs 100k - 1 million L2 car.

 

[Doggydogworld]

Finally, do the statistics count accidents where the AP was not "engaged" at the point of impact because the driver had applied the brakes or turned the wheel to try and avoid the accident like an attentive driver should? We don't know because Tesla doesn't tell us how they came up with these.

But Tesla does tell us: "To ensure our statistics are conservative, we count any crash in which Autopilot was deactivated within 5 seconds before a crash"

My point is he was well aware the car could disengage at anytime, but decided to watch a video on a portable DVD player anyways. Knowing the specific about not designed for cross traffic is irrelevant, as he was not looking at the road anyways.

"He was watching Harry Potter" is a myth, reported by some media but contradicted in the NTSB report.

 

[Soda Popinski]

"He was watching Harry Potter" is a myth, reported by some media but contradicted in the NTSB report.

Interesting. Based on an article reporting the crash, it was the truck driver that stated the movie was playing. I could see why he'd want to put the blame on the Tesla driver (Truck driver failed to yield to oncoming traffic). Maybe he was listening to the soundtrack. The NTSB report did mention 2 laptops on the scene, as from the article, cop at the scene mentioned a DVD player, which could easily have been one of the laptops mis-identified.

 

[mikes_fsd]

~40+ million L2 cars

And not a single real-world stat to show for it! That's how well used it is!