March of 9s: The Beginning of the End

I have had enough experience now with FSD 12 to say that in my opinion, I believe Ashok is correct:

https://twitter.com/x/status/1774673326583210476

I would also stress we are at the beginning of the end. That is, we will hopefully now start to see significant improvement with the "known FSD issues" and "corner case disengagements" with every major FSD release. "The end" (level 3 /4 autonomy within a wide Operational Design Domain (ODD)) is almost certainly year+ away. I'd like to document, just from one consistent anecdotal use case that can be repeated over time, where we are "starting from."

I will be driving a 90ish mile (2-3 hrs) "loop" under FSD to cover a range of driving scenarios. (Mileage is approximate):

For privacy reasons, there are about 15 miles in my loop that are not included in the link above that takes me to/from my actual home.

Here is a Link to inspect the route in detail

This route takes me from NJ Suburbs into and out of Manhattan (NYC) and includes approximately:

• 10 Miles Suburban driving
• 65 Miles "limited access highway" driving (This should be using the FSD 'highway stack' which is not the same as the FSD 12 stack)
  • includes interchanges
  • Includes Tolls
  • Includes Tunnel
• 8 miles of other "highway type driving", (will probably fall under the FSD 12 stack)
• 6 miles of dense city driving...including areas around Times Square, Rockefeller center, etc which will have dense vehicle and pedestrian traffic.

I will not be recording with a phone or anything like that. However I will try and save dashcam footage of anything notable.

I will report on:

1. Interventions (Accelerator presses, particularly if safety related)
2. Disengagements (comfort or safety related)
3. Overall impressions

As we know Version 12.3.x does not support...(but will need to in the future):

1. Smart Summon or "Banish"..so what I call the "first 100 yards and final 100 yards" is not available to test. (Drop offs / pick-ups).
2. "Reversing" while on FSD is not yet supported

Finally, there are what I would say 2 well documented "comfort / safety" issues with FSD 12.3.x that I have also experienced regularly first hand:

1. "Lane Selection Wobble"...for example, approaching an intersection where the single driving lane splits into multiple lanes (turning vs. straight)...FSD may act "indecisively"
2. Unprotected turn (stop sign) behavior. Notably: stops early....then creeps. If no cars detected it may creeps into intersection instead of "just going". Further, if it has creeped into the intersection, THEN detects a car approaching, it may still hesitate and require intervention (accelerator press) to get it going.

In addition to those two consistent issues, I expect to encounter some issues related to routing, and any number of other 'corner case' issues. All things that will ultimately need to be handled, but we expect to see dealt with as we progress though the "March of 9s"...toward the "end of the end".

Although I have driven FSD regularly over the past 3 weeks...I have yet to take it into NYC.

Vehicle: Refresh Model S (2023), Vision Only, HW4. First test will be using:
Firmware version: 11.1 (2024.3.15)
FSD Version: 12.3.4

So...there's the set-up. I expect later today to drive the first loop.

 

[kabin]

To be clear...According to Tesla, a human driving on Supervised FSD has been safer than a human driving without FSD for years. Tesla has never claimed that unsupervised FSD is safer than humans....yet.

It doesn't make much sense to say a human driver is safer than a human driver. You can't have it both ways.

I should add: on the other hand, if FSD performance was comparable to an attentive human driver we could begin to argue supervised FSD safety approaches that of a human attentive driver minus FSD. Until then it's just a driver convenience needing kid gloves.

 

[Joey D]

It doesn't make much sense to say a human driver is safer than a human driver. You can't have it both ways.

It's not difficult a concept:

A self-driving car, being supervised by a human driver...is safer than a car being driven by a human driver.

 

[Joey D]

One of my places to visit!
Unfortunately, L5 cars aren't at liberty to make their share of mistakes.

Actually...they are. They do not have to be perfect.

 

[Dewg]

Actually...they are. They do not have to be perfect.

Thought experiment: how long would the company behind that car survive while their cars aren't driving perfectly?

 

[kabin]

It's not difficult a concept:

A self-driving car, being supervised by a human driver...is safer than a car being driven by a human driver.

At best it's similar to a teenage permit driver with parental oversight to that of a human driver. There is no safety advantage over the latter.

 

[Joey D]

Thought experiment: how long would the company behind that car survive while their cars aren't driving perfectly?

That depends on how much better their cars are than human drivers. What matters is how much better they are....not that they are perfect. Perfection is impossible.

We get into cabs and ubers all the time even though the drivers are not perfect.

 

[Joey D]

At best it's similar to a teenage permit driver with parental oversight to that of a human driver. There is no safety advantage over the latter.

Flawed analogy. The safety advantage depends on how good the "teenage permit driver is."

Put another way:
I can supervise Michael Schumacher as the driver...
Or I can supervise a blind 90 year old as the driver....

I'd say there is a safety difference even though the same person is doing the supervising. And if I'm "driving" on a racetrak....the combination of Schumacher and me as a supervisor...is probably safer than me driving alone.

 

[uscbucsfan]

That depends on how much better their cars are than human drivers. What matters is how much better they are....not that they are perfect. Perfection is impossible.

We get into cabs and ubers all the time even though the drivers are not perfect.

That's how things would work in a rational world, but not in reality. A small amount of incidents will have local jurisdictions and NHTSA looking into a company and potentially banning ADAS in that area.

That's reality. The cars will need to be nearly perfect, unfortunately.

 

[Joey D]

That's how things would work in a rational world, but not in reality. A small amount of incidents will have local jurisdictions and NHTSA looking into a company and potentially banning ADAS in that area.

That's reality. The cars will need to be nearly perfect, unfortunately.

I actually disagree.

I see it more like air flight.

Accidents happen. When they do, they will be investigated and cause determined, and corrections will be made...and it will get that much safer. There will be continuous improvement. The FAA may temporarily halt flights or a model of aircraft after an incident, but as long as a correctable issue is resolved and risk is minimized, we carry on.

 

[spacecoin]

It's not difficult a concept:

A self-driving car, being supervised by a human driver...is safer than a car being driven by a human driver.

You clearly do not understand anything about AVS if you believe this. Go back and look at the diagram I posted?

The human + system is only as safe as a human under two conditions:
1) The system is so bad the human isn't trusting the system
2) The system is so good it by itself is better than a human.

One word: automation complacency.

 

[Joey D]

One word: automation complacency.

That's two.

 

[spacecoin]

That's two.

Fine. You still don't understand what you're talking about.

 

[uscbucsfan]

I actually disagree.

I see it more like air flight.

Accidents happen. When they do, they will be investigated and cause determined, and corrections will be made...and it will get that much safer. There will be continuous improvement. The FAA may temporarily halt flights or a model of aircraft after an incident, but as long as a correctable issue is resolved and risk is minimized, we carry on.

Real world data says otherwise.

If cars are blocking roads or causing accidents, local governments will ban them. Unlike air travel, this is mostly handled by individual states and cities, so there's not some large federal agency (yet) dictating or enabling these.

There will be cities that ban them at the drop of a hat, politically or due to "concern of local citizens" as we saw before.

 

[spacecoin]

Real world data says otherwise.

If cars are blocking roads or causing accidents, local governments will ban them. Unlike air travel, this is mostly handled by individual states and cities, so there's not some large federal agency (yet) dictating or enabling these.

There will be cities that ban them at the drop of a hat, politically or due to "concern of local citizens" as we saw before.

100% agree. That's why the companies need to work with the communities where you aim to deploy and buy some slack, like Waymo has been doing over the last 10 years. Educating first responders, community leaders and politicians. Its a trust business.

 

[uscbucsfan]

100% agree. That's why the companies need to work with the communities where you aim to deploy and buy some slack, like Waymo has been doing over the last 10 years.

Yet, Waymo has had suspensions from cities as well.

It's inevitable. Which is why at the beginning, it's going to be a battle and costly for Tesla, just like it is for Waymo, Zoox, etc.

Just less costly because the cars sensors cost less.

 

[Joey D]

Real world data says otherwise.

If cars are blocking roads or causing accidents, local governments will ban them. Unlike air travel, this is mostly handled by individual states and cities, so there's not some large federal agency (yet) dictating or enabling these.

There will be cities that ban them at the drop of a hat, politically or due to "concern of local citizens" as we saw before.

We will have to agree to disagree.

The sheer convenience and cost benefits enabled by automated driving puts immense political pressure on the other side.

People die in airlines. But the benefit of air flight means that we as a society mitigate risks...but do not eliminate flying.

 

[spacecoin]

Yet, Waymo has had suspensions from cities as well.

It's inevitable. Which is why at the beginning, it's going to be a battle and costly for Tesla, just like it is for Waymo, Zoox, etc.

Just less costly because the cars sensors cost less.

Tesla isn't really in that business at all, and Waymo is years ahead of Cruise and Zoox in all aspects.

 

[uscbucsfan]

We will have to agree to disagree.

The sheer convenience and cost benefits enabled by automated driving puts immense political pressure on the other side.

People die in airlines. But the benefit of air flight means that we as a society mitigate risks...but do not eliminate flying.

Not sure it's debatable. San Francisco, San Mateo, and others have suspended driverless taxis after relatively minor incidents in the past (also some major incidents)...now imagine somewhere like Columbia, South Carolina.

Most people aren't logical...nor are governments. We see regulators and news agencies hunting for AP wrecks, ignoring non-AP wrecks, because "the person wasn't driving". Tech is more scrutinized than people. It's reality.

 

[Joey D]

Not sure it's debatable. San Francisco, San Mateo, and others have suspended driverless taxis after relatively minor incidents in the past (also some major incidents)...now imagine somewhere like Columbia, South Carolina.

How are automated cars operating in San Fran if automated driving has been suspended?

Most people aren't logical...nor are governments.

No argument there. But people do relentlessly demand "cheaper" , "more convenient", and "enabling." Under enough popular pressure, governments cave.

 

[uscbucsfan]

How are automated cars operating in San Fran if automated driving has been suspended?

Again, they resumed them, but now imagine a place that is much less tolerant to "robots" or EVs.

Every incident in places like California they have a town hall and debate the merits of them operating.

As I said in the investor thread, RT needs to be significantly cheaper than Uber to take market share. There are many who dislike EVs and ADAS out of being extremist idiots.