FSD V GM CRUISE

[Daniel in SD]

Sorry, that's an absurd definition. Capabilities are defined by what a system can do, not what you hope it can do. It's like calling a medical student a neurosurgeon because that's what they're planing on doing. It also goes against what you were arguing in this thread.

I have to ask, though, where did you get this quote? When I search the SAE site, I couldn't find it anywhere.

It's from the actual SAE document (SAE J3016_202104.pdf - UNECE Wikihttps://wiki.unece.org › download › attachments). I think the intent is to prevent companies from circumventing autonomous vehicle testing regulations by claiming their prototypes are driver assist features. Back in 2016 Uber tried to claim their robotaxi prototypes were L2 and got shut down by the CA DMV.

This is a more rational definition and contradicts the quote you posted.

That quote is just saying that when the Mercedes S-class has its L3 system enabled it's L3 and when it has its L2 system enabled it's L2.

 

[Daniel in SD]

So what happens when you need an extra 200,000 or more cars at once? Cat 4 Hurricane headed toward Louisiana and Mississippi. You need to evacuate 3 million people.

Isn't there usually days of warning for hurricanes? You can do multiple round trips.

 

[sleepydoc]

So what happens when you need an extra 200,000 or more cars at once? Cat 4 Hurricane headed toward Louisiana and Mississippi. You need to evacuate 3 million people.

Well, from the TV pictures I've seen all the cars are sitting still on the highway anyway, so it doesn't matter!

 

[sleepydoc]

It's from the actual SAE document (SAE J3016_202104.pdf - UNECE Wikihttps://wiki.unece.org › download › attachments). I think the intent is to prevent companies from circumventing autonomous vehicle testing regulations by claiming their prototypes are driver assist features. Back in 2016 Uber tried to claim their robotaxi prototypes were L2 and got shut down by the CA DMV.


That quote is just saying that when the Mercedes S-class has its L3 system enabled it's L3 and when it has its L2 system enabled it's L2.

Right, but you were making a big point of saying Mercedes beat Tesla to Level 3, but now you're making a point of saying that a system is defined not by what it can do but by what the ultimate goal is and by that measure Tesla beat everyone. You can't have it both ways.

Reading the text again:

The level of a driving automation system feature corresponds to the feature’s production design intent. This applies regardless of whether the vehicle on which it is equipped is a production vehicle already deployed in commerce, or a test vehicle that has yet to be deployed. As such, it is incorrect to classify a Level 4 design-intended ADS feature equipped on a test vehicle as Level 2 simply because on-road testing requires a test driver to supervise the feature while engaged, and to intervene if necessary to maintain operation.

it seems they intend this to apply to a fully functional system that's being tested; Tesla's software, as designed now requires driver intervention (via the steering wheel) so one can make a strong argument that the intent is not a level 3/4/5 system but a level 2 system. I'm sure this is what the debate you referenced was about.

Regardless, Defining a system that is under development and is still limited in capabilities as a higher level based on future goals rather than its current capabilities is still absurd.

 

[AndrewZ]

In theory once robotaxis are ubiquitous, and enough time goes by (one decade, two?) nobody needs to own a car, you just summon one, and the fleets of these robotaxis would pre-position based on historical demand to keep summon time low.

Yes, there will be us that like to drive, and we will be able to do so as long as the Govt allows it, perhaps pushing us out to the track eventually for entertainment like ATV riding.

The rest of the road will be public transport via robotaxi. People getting their drivers license will be a thing of the past.

Welcome to a possible future...

This is a pretty interesting vision of a distant future, but I can immediately and definitively say that some people/companies will 100% still need to own individual vehicles. This type of setup might make sense for urban centers and areas that are densely populated, and these places likely already see vehicle ownership rates far lower than national averages with much higher usage of stuff like Uber, but it does not make sense for areas that are rural or less densely populated. And I think any government mandate, like only allowing vehicle ownership among people who have rural addresses or something, would not be realistically enforceable.

There are tons of reasons for individual vehicle ownership besides just enjoyment, but I can see something like what you're describing happening in urban centers within big cities.

 

[Daniel in SD]

Right, but you were making a big point of saying Mercedes beat Tesla to Level 3, but now you're making a point of saying that a system is defined not by what it can do but by what the ultimate goal is and by that measure Tesla beat everyone. You can't have it both ways.

In a sense Tesla has beat everyone to L5. haha. FSD Beta could be an L5 system that's still in development (Tesla claims it's not! I say it is but those arguments get nowhere). Mercedes is deploying an L3 system, meaning it will no longer requires a safety driver.

it seems they intend this to apply to a fully functional system that's being tested; Tesla's software, as designed now requires driver intervention (via the steering wheel) so one can make a strong argument that the intent is not a level 3/4/5 system but a level 2 system. I'm sure this is what the debate you referenced was about.

Regardless, Defining a system that is under development and is still limited in capabilities as a higher level based on future goals rather than its current capabilities is still absurd.

Well I disagree. All the major self-driving car companies have sophisticated safety driver monitoring. I don't see anything magical about touching the steering wheel occasionally, especially when it's not even being used as a control. Basically you're saying that testing of self-driving vehicles with a safety driver present should not be regulated. Fair enough, maybe there shouldn't be regulations. FSD Beta seems to be safe enough.

 

[AndrewZ]

I don't even see Level 5 and FSD being spoken about in the same breath anymore, the narrative has switched to Level 4 functionality. Don't know when it happened or how everyone seemed to have changed their minds simultaneously, but nowadays Elon is only ever asked about achieving Level 4.

True Level 5 autonomy seems to have been relegated to the fantasy realm for now

 

[DJT1]

This is a pretty interesting vision of a distant future, but I can immediately and definitively say that some people/companies will 100% still need to own individual vehicles. This type of setup might make sense for urban centers and areas that are densely populated, and these places likely already see vehicle ownership rates far lower than national averages with much higher usage of stuff like Uber, but it does not make sense for areas that are rural or less densely populated. And I think any government mandate, like only allowing vehicle ownership among people who have rural addresses or something, would not be realistically enforceable.

There are tons of reasons for individual vehicle ownership besides just enjoyment, but I can see something like what you're describing happening in urban centers within big cities.

Yes, it would initially probably work like car ownership in Tokyo today. If you can’t prove you have a garage space for a car, you can’t get (register) a car. In the future in urban centers if you don’t have a garage spot, you will be limited to robotaxis or other forms of ppublic transport.

 

[AlanSubie4Life]

Why accelerate the build out of the supercharger network with chargers that require hands to plug into the car, rather than some automated plug-in design.

This is obviously something Tesla has been thinking about, and for interstate robotaxi travel they have likely decided that a couple of Optimus attendants per Supercharger station is cheaper than retrofitting every single pedestal. It makes sense to wait on robotaxis to be finalized since then Optimus can use a branch of that code. He’ll have a hut next to the charging cabinets.

 

[Bitdepth]

Right, but you were making a big point of saying Mercedes beat Tesla to Level 3, but now you're making a point of saying that a system is defined not by what it can do but by what the ultimate goal is and by that measure Tesla beat everyone. You can't have it both ways.

Note: Design Intention is not defined as hope or future capability or goal. Design intention means the specifications and capabilities as defined by the manufacturer of the system.

Page 36

The level assignment rather expresses the design intention for the feature and as such tells potential users or other interested parties that the feature can be expected to function such that the roles of the user versus the driving automation system while the feature is engaged are consistent with the assigned level, as defined in this document. The level assignment is typically based on the manufacturer’s knowledge of the feature’s/system’s design, development, and testing, which inform the level assignment. An ADS feature’s capabilities and limitations are expected to be communicated to prospective users through various means, such as in an owner’s manual, which explains the feature in more detail, such as how it should and should not be used, what limitations exist (if any), and what to do (if anything) in the event of a DDT performance-relevant system failure in the driving automation system or vehicle. As such, the manifestation of one or more performance deficiencies in either the driving automation system or in the user’s use of it does not automatically change the level assignment. For example

Mercedes is actually deploying a L3 system which is also their design intention for that particular system. Same vehicle can have a L4, L3 and L2 modes. Note how they do not claim the vehicle L3 mode is intended to do L4 someday in the future. Mercedes differentiates those modes and their capabilities.

Mercedes-Benz – the front runner in automated driving and safety technologies | Mercedes-Benz Group

Since the invention of the automobile in 1886, Mercedes-Benz has repeatedly set standards in vehicle safety with new innovative technologies. Safety is a core element of the Mercedes-Benz brand.

group.mercedes-benz.com

L2 Mode Intelligent Drive
(Does the usual lane keep, lane change, adaptive cruise control with constant human supervision)

L3 Mode DRIVE PILOT
(Does the lane keep and adaptive cruise control on specific sections of divided highway below 30mph without human supervision but may be requested at any time)

L4 Mode INTELLIGENT PARK PILOT²
(Parks cars without human supervision in specific garages equipped for it.)

Regardless, Defining a system that is under development and is still limited in capabilities as a higher level based on future goals rather than its current capabilities is still absurd.

Design intent does not mean future goals. It means the capability of the system as described by the manufacturer of said system. If Tesla says FSD beta is L5 then it is. It would be a bad L5 system with the current issues it has, but it will be treated as L5 and when it gets into an accident Tesla bares full responsibility and liability for any damages as a result of the accident. Uber learned that the hard way.

 

[Daniel in SD]

This is obviously something Tesla has been thinking about, and for interstate robotaxi travel they have likely decided that a couple of Optimus attendants per Supercharger station is cheaper than retrofitting every single pedestal. It makes sense to wait on Robotaxis to be finalized since then Optimus can use a branch of that code. He’ll have a hut next to the charging cabinets.

I had forgotten that the Optimus prototype will be shown later this year. I admit that I was stumped about how the cross country summon would work. Hopefully Optimus will also be able to move my legacy HW2.5 car after I finish charging so I can avoid idle fees.

It's funny that making a humanoid robot that can operate a Supercharger is probably a much easier engineering problem than making a driverless car. For one thing the required MTBF is many orders of magnitude less.

 

[Dewg]

I've always hoped for wireless charging for the car, though I think the current is not high enough for fast charging. I remember seeing some "pads" in garages years ago, but haven't seen them since.

 

[Daniel in SD]

I've always hoped for wireless charging for the car, though I think the current is not high enough for fast charging. I remember seeing some "pads" in garages years ago, but haven't seen them since.

At best they're 90% efficient and they require AC. So now to do 250kW you're dissipating 25kW of heat somehow and you've got to have a massive AC to DC converter in the car (the current one is only 12kW). For reference look at the size of the supercharger cabinets.

 

[Supcom]

I had forgotten that the Optimus prototype will be shown later this year. I admit that I was stumped about how the cross country summon would work. Hopefully Optimus will also be able to move my legacy HW2.5 car after I finish charging so I can avoid idle fees.

It should be fairly straightforward to train Optimus to call for a car hauler and load the car for you.

 

[sleepydoc]

At best they're 90% efficient and they require AC. So now to do 250kW you're dissipating 25kW of heat somehow and you've got to have a massive AC to DC converter in the car (the current one is only 12kW). For reference look at the size of the supercharger cabinets.

Even 90% is mostly a pipe dream - the MagSafe chargers for iPhones are pretty optimized for perfect alignment, etc and only have 75% efficiency. Regardless, part of the reason for driving an EV is energy conservation. It would be incredibly counter productive to needlessly waste 10-25% of the energy with a wireless charging system.

 

[sleepydoc]

Note: Design Intention is not defined as hope or future capability or goal. Design intention means the specifications and capabilities as defined by the manufacturer of the system.

Page 36


Mercedes is actually deploying a L3 system which is also their design intention for that particular system. Same vehicle can have a L4, L3 and L2 modes. Note how they do not claim the vehicle L3 mode is intended to do L4 someday in the future. Mercedes differentiates those modes and their capabilities.

Mercedes-Benz – the front runner in automated driving and safety technologies | Mercedes-Benz Group

Since the invention of the automobile in 1886, Mercedes-Benz has repeatedly set standards in vehicle safety with new innovative technologies. Safety is a core element of the Mercedes-Benz brand.

group.mercedes-benz.com

L2 Mode Intelligent Drive
(Does the usual lane keep, lane change, adaptive cruise control with constant human supervision)

L3 Mode DRIVE PILOT
(Does the lane keep and adaptive cruise control on specific sections of divided highway below 30mph without human supervision but may be requested at any time)

L4 Mode INTELLIGENT PARK PILOT²
(Parks cars without human supervision in specific garages equipped for it.)


Design intent does not mean future goals. It means the capability of the system as described by the manufacturer of said system. If Tesla says FSD beta is L5 then it is. It would be a bad L5 system with the current issues it has, but it will be treated as L5 and when it gets into an accident Tesla bares full responsibility and liability for any damages as a result of the accident. Uber learned that the hard way.

Yeah, I agree, but @Daniel in SD was arguing otherwise.

 

[Dewg]

Even 90% is mostly a pipe dream - the MagSafe chargers for iPhones are pretty optimized for perfect alignment, etc and only have 75% efficiency. Regardless, part of the reason for driving an EV is energy conservation. It would be incredibly counter productive to needlessly waste 10-25% of the energy with a wireless charging system.

Ugh - that's terrible waste. If they can achieve 94%, it might be a good solution for trickle charging. For those that commute 20-40 miles a day, the wireless charger could keep you topped off overnight, or during the 8 hours at work. But I agree with you that below 90%, it doesn't seem worth it.

I also read years ago about a charging system built into the roads themselves - obviously a very optimistic future endeavor, where there are alternating magnetic strips placed in the middle of a travel lane on a freeway/highway, and some sort of flywheel under the car that rotates as it passes over the magnets, generating a trickle charge. It wouldn't likely give you a significant charge, but could increase efficiency of your car, perhaps 10-20% additional range.

 

[ThomasD]

Yes, it would initially probably work like car ownership in Tokyo today. If you can’t prove you have a garage space for a car, you can’t get (register) a car. In the future in urban centers if you don’t have a garage spot, you will be limited to robotaxis or other forms of ppublic transport.

So carports and driveways are out

 

[Dewg]

So carports and driveways are out

Or my neighbors who fill their garages to capacity with "stuff", as George Carlin used to call it. I have one neighbor who installed his Tesla charger just inside his garage door and parks outside (since the garage is full) with the charging cable going under the door and into his cars (he has two Model Ys and I just saw two weeks ago a new Rivian in his driveway - and they just installed solar today).

 

[ThomasD]

Would magnets in the road collect metal debris? So how long before urban areas in every state will have robotaxis? Will urban areas that have very little public transit or even no public transit get robotaxis?