Here's How Tesla will launch the 'Tesla Network' as L2 ADAS!

[Bladerskb]

So Elon might pull off another hat trick. I have always considered the initial launch of AP1 to be dangerous only because Tesla allowed up to 30 mins of no nags and put absolutely no restrictions on it. Now they are learning but the 4 deaths that have happened with AP could have been avoided. But since actual full autonomy isn't here and won't be here for a while, although Elon has been saying its "two years away" since 2015.

Elon will no doubt continue to deceive and say that FSD is here. But since he can't deliver, he will instead will do 'FSD' as a L2 system. I have believed this before but now i know i'm right Looks like i was right. Elon will try to launch "Tesla Network" as L2 ADAS. It will be a uber-like Bird/Limes with pick up spots in downtowns and parking lots. The driver being responsible.

It will be similar to AP but for streets. It will require a driver and require the driver to pay attention to supervise the "FSD" system. Here's why Its brazen, brilliant & dangerous but might actually work.

Its brazen because it relies on untrained drivers who ain't familiar with Teslas to be attentive. Its brilliant because it solves the problem of waiting for L5 (L4 everywhere), increases mind-share, adds revenue, app market-share, but continues myth that Teslas are actually self-driving and are the leader. Dangerous because people will die as result.

It will be a Uber-like Bird/Limes/MoGo. There will be two ride options, miles base and time(hours) based.
There will pickup spots which are also drop off spots around downtown areas and parking lots in downtown around the country.

Miles Based Option:

You go to a pick up spot, use your phone and request a ride to your destination, it will select a drop off spot closest to your destination, it will assign you a car which you get in.
And it will drive with you as the driver and will alert you to be fully prepared to take over at any time that you have to accept. They will also use the rear mirror camera to watch you to make sure you don't fall asleep or do anything crazy (would be crazy if they just relied on the wacky steering sensor). Once you get to the pick-up spot that is closet to your destination, it will park and you get out and go about your day.

Time Base Option:

The car will take you to any destination with you supervising. You get out, do whatever you want to do and then when its time to drop the car back, you get in and have it navigate to the closest pickup/drop off.

This is similar to electric scooters bikes that are based on hours(time) based not miles based, so you are allotted say 6 hours and you can take it home, run errands, etc.

This can actually work and allow Elon to still claim FSD, winning mind-shares as people will continue to believe Teslas are self driving, while also making money. It will technically still be "Tesla Network". But there will be casualties.

 

[electronblue]

This is not a bad theory, they certainly might try something like that — anything to save face I guess.

 

[WarpedOne]

My sh*tl*st just got 2 more items.

 

[S4WRXTTCS]

Possibly, but I think there is one big miss with this.

I think we all want Grand Theft Auto style freedom when it comes to driving.

Why can't I just get in a Tesla, and drive it into the city where it then parks itself?

That way I didn't have to go through the absolute worst part of going to the city.

I don't care how they pull off auto-parking, but they absolutely need auto-parking to work in order to really sell people on the Tesla network the way you described.

90% of the fun of something like Bird/Lime is you can just go anywhere without having to worry too much about where you leave them.

As to it being dangerous? Absolutely.

But, when it comes down to it I'd say it's convenience versus safety.

Lots of people have injured/killed themselves on electric scooters, and ride share bikes. They do because they ride them dangerously and they don't wear a helmet.

The first 2 AP deaths were people who obviously weren't paying attention. They made the conscious decision to ignore the warnings.

The 3rd one I have a lot more sympathy for as there likely wasn't much notice. Plus he had the terrible luck of crashing in the exact same spot that another vehicle crash just days before, and the crash barrier wasn't reset. Lots of things went wrong that led to his death.

The 4th one we don't know if he was on AP or not. We don't have enough details, but we do know there likely was a lot of advanced notice given how far the semi made it across the road before the accident happened.. We also know that very little time is given until the system nags you. It wasn't like the old days of AP1. My bet is the driver simply fell asleep as it was really early in the morning.

We also don't know how many people AP saved from momentary inattentions. Humans don't have redundancy so any time a sneeze happens, or anytime some idiot text is a time where a little backup can come in handy.

Plus there are lots of people who likely have anxieties towards driving that are attracted to Tesla's.

In fact I readily admit if Tesla pulls off any kind of L2 FSD that's fairly decent that I'l buy my mom one. Why? Cause I can't get her to give up her keys, and she's been stubbornly opposed to lyft/uber. She's still mentally there so I trust her in that she won't drive when she's physically not in a state that's safe. In the mean time she has Subaru with eyesight, and that I find somewhat comforting as it's not a bad system for the mostly city driving she does.

 

[electronblue]

Possibly, but I think there is one big miss with this.

I think we all want Grand Theft Auto style freedom when it comes to driving.

Why can't I just get in a Tesla, and drive it into the city where it then parks itself?

That way I didn't have to go through the absolute worst part of going to the city.

I don't care how they pull off auto-parking, but they absolutely need auto-parking to work in order to really sell people on the Tesla network the way you described.

90% of the fun of something like Bird/Lime is you can just go anywhere without having to worry too much about where you leave them.

I hope we learn more on the 22nd but in the meanwhile I guess it is useful to consider what is realistic.

Tesla is now saying it will keep leased Model 3s after the lease for the Network, so I guess after 3 years perhaps it will start accumulating a fleet of Model 3s for this purpose. Of course if they are not ready they could just sell them as CPOs. This could be just a marketing ploy and a way to avoid committing to a residual. But assuming they are serious, would they have Level 4 ready in three years with the current sensor suite, including car responsible parking maneuvers without a nose camera?

What @Bladerskb described is in my view the first really serious attempt at answering how this could be done before Level 4 is widely ready so at the very least it is novel speculation and thus appreciated.

 

[S4WRXTTCS]

I hope we learn more on the 22nd but in the meanwhile I guess it is useful to consider what is realistic.

Tesla is now saying it will keep leased Model 3s after the lease for the Network, so I guess after 3 years perhaps it will start accumulating a fleet of Model 3s for this purpose. Of course if they are not ready they could just sell them as CPOs. This could be just a marketing ploy and a way to avoid committing to a residual. But assuming they are serious, would they have Level 4 ready in three years with the current sensor suite, including car responsible parking maneuvers without a nose camera?

What @Bladerskb described is in my view the first really serious attempt at answering how this could be done before Level 4 is widely ready so at the very least it is novel speculation and thus appreciated.

Before Blader mentioned this approach I was of the opinion that the Tesla Network was just a ploy to push up the stock market. Investment people love betting on a company on things way down the road.

I'm still not sure if it's realistic.

If the vehicles had 360 degree down facing cameras for low speed driving in crowded areas I'd say yes. It would also help if they had rear corner radars.

It's hard for me to envision and kind of FSD due to visibility issues when it rains. Like today I was getting the dancing cars in the visualization, and I believe it was due to the rain. A little while later after the sensors heated up/dried out I didn't have any issue with the dancing cars.

I look through the rear camera while its raining and there is no way I can identify a car being behind me. It's just a blur.

I'm still skeptical. I think it's going to require at least one sensor suite revision.

In any case I like that Blader placed a bet on something fairly near term.

That I have to give him credit for.

Oh, and I'm also really happy that Tesla has finally included AP as standard. So it could be that I'm just in a happy mood.

 

[S4WRXTTCS]

One the 22nd I'm hoping we'll hear more details about Tesla insurance.

I don't see any of this working unless there is insurance coverage for cars involved in the network.

Like lets say I bought my mom a FSD Model 3.

I'd want her to share the car on the Tesla network as she doesn't drive that much. Obviously her insurance would nix that plan. So she'd have to get some insurance that would allow it.

The only way I see it working is if Tesla partnered up with some insurance company for it.

 

[electronblue]

@S4WRXTTCS

I don’t disagree with anything you say there.

One more thought: What if you don’t think about Tesla Network so much as an autonomous driving network but as a built-in geofence and billing network. In this speculation the car would allow and indeed require limited driving from the driver but at the same time control where the car goes and how much is billed for it. Weather would not matter that much here, driver could just take over as needed but the car would refuse to stray too far.

I could actually see Tesla lauching this feature to owners in a not too distant future. You can place your car in the network, Tesla takes a share, people drive your semi-autonomous car and the software keeps you car relatively safe and the money flowing. It is not bad speculation.

 

[S4WRXTTCS]

@S4WRXTTCS

I don’t disagree with anything you say there.

One more thought: What if you don’t think about Tesla Network so much as an autonomous driving network but as a built-in geofence and billing network. In this speculation the car would allow and indeed require limited driving from the driver but at the same time control where the car goes and how much is billed for it. Weather would not matter that much here, driver could just take over as needed but the car would refuse to stray too far.

I could actually see Tesla lauching this feature to owners in a not too distant future. You can place your car in the network, Tesla takes a share, people drive your semi-autonomous car and the software keeps you car relatively safe and the money flowing. It is not bad speculation.

I don't disagree with any of that as it is plausible.

But, I know human nature is to be lazy.

Having self-parking of even a limited kind would be absolutely killer.

Where Tesla partnered up with parking places around a city, and geo-fenced areas where it could park itself.

There used to be a valet service that would park a persons Tesla at a charger. All you had to do is drive close to where the valet was, and they'd do the rest. They disappeared from the Seattle area before I had the chance to try it.

 

[diplomat33]

So Elon might pull off another hat trick. I have always considered the initial launch of AP1 to be dangerous only because Tesla allowed up to 30 mins of no nags and put absolutely no restrictions on it. Now they are learning but the 4 deaths that have happened with AP could have been avoided. But since actual full autonomy isn't here and won't be here for a while, although Elon has been saying its "two years away" since 2015.

Elon will no doubt continue to deceive and say that FSD is here. But since he can't deliver, he will instead will do 'FSD' as a L2 system. I have believed this before but now i know i'm right Looks like i was right. Elon will try to launch "Tesla Network" as L2 ADAS. It will be a uber-like Bird/Limes with pick up spots in downtowns and parking lots. The driver being responsible.

It will be similar to AP but for streets. It will require a driver and require the driver to pay attention to supervise the "FSD" system. Here's why Its brazen, brilliant & dangerous but might actually work.

Its brazen because it relies on untrained drivers who ain't familiar with Teslas to be attentive. Its brilliant because it solves the problem of waiting for L5 (L4 everywhere), increases mind-share, adds revenue, app market-share, but continues myth that Teslas are actually self-driving and are the leader. Dangerous because people will die as result.

It will be a Uber-like Bird/Limes/MoGo. There will be two ride options, miles base and time(hours) based.
There will pickup spots which are also drop off spots around downtown areas and parking lots in downtown around the country.

Miles Based Option:

You go to a pick up spot, use your phone and request a ride to your destination, it will select a drop off spot closest to your destination, it will assign you a car which you get in.
And it will drive with you as the driver and will alert you to be fully prepared to take over at any time that you have to accept. They will also use the rear mirror camera to watch you to make sure you don't fall asleep or do anything crazy (would be crazy if they just relied on the wacky steering sensor). Once you get to the pick-up spot that is closet to your destination, it will park and you get out and go about your day.

Time Base Option:

The car will take you to any destination with you supervising. You get out, do whatever you want to do and then when its time to drop the car back, you get in and have it navigate to the closest pickup/drop off.

This is similar to electric scooters bikes that are based on hours(time) based not miles based, so you are allotted say 6 hours and you can take it home, run errands, etc.

This can actually work and allow Elon to still claim FSD, winning mind-shares as people will continue to believe Teslas are self driving, while also making money. It will technically still be "Tesla Network". But there will be casualties.

First of all, it's a theory. I don't know how you can present a theory as a "fact" and say you that you know you are right. You did not present any specific evidence, you just laid out the details of your theory.

Second, I see two ways where you might be wrong:
1) It's too dangerous. There is no way you are going to let untrained folks who don't know Teslas supervise the car's FSD. I've given Model 3 rides to my friends and some of them who have never seen a Tesla before, did not even know how to work the door handle on the Model 3!!! I watched them fiddle with the door handle trying to figure it out! You think an owner or Tesla itself is going to leave these folks responsible for supervising FSD?!?! No way! Frankly, that's not sustainable. As soon as accidents start happening or customers have bad experiences, it would be the death of the Tesla Network.

The only way it could maybe work is if Tesla's autonomy is good enough that very little intervention is actually required. Or it could work if Tesla hires safety drivers for the Tesla Network.

2) Tesla continues to actually claim full autonomy is coming. Tesla actually puts in the Model 3 lease:

"Please note, customers who choose leasing over owning will not have the option to purchase their car at the end of the lease, because with full autonomy coming in the future via an over-the-air software update, we plan to use those vehicles in the Tesla ride-hailing network"
Tesla launches Model 3 leases, will keep cars for autonomous Uber-like service after term

How does this square with your theory? So Tesla will use Model 3's after the lease for the Tesla Network. They are promising "full autonomy" but will require strangers to monitor that full autonomy? That seems very unlikely.

Here's my theory: Tesla will hire Tesla employees to be safety drivers. So the cars will drive around autonomously and pick people up but with a Tesla employee as a safety driver.

 

[electronblue]

Here's my theory: Tesla will hire Tesla employees to be safety drivers. So the cars will drive around autonomously and pick people up but with a Tesla employee as a safety driver.

That is not impossible either but I would consider that a variant of what @Bladerskb is suggesting unless they want to call the system Level 4 with all the responsibility that entails.

As for the ”full autonomy” quote. In Tesla speak it is not that simple. They might strive for full autonomy while still in actual fact releasing a Level 2 ADAS system at this time. Not unlike Elon calling NoA as FSD on highways...

 

[diplomat33]

That is not impossible either but I would consider that a variant of what @Bladerskb is suggesting unless they want to call the system Level 4 with all the responsibility that entails.

Yes, it's a variation on @Bladerskb theory but it has one major difference. Having Tesla employees who are qualified and trained, act as safety drivers is a huge difference from letting untrained strangers act as safety drivers. My theory removes one of the key concerns that blader raises (ie the risk of having untrained strangers supervise the car).

But the advantage that blader has is even if my variation is correct, he will still say he is right even though it solves one of his concerns. if Tesla has any form of driver attention requirement during autonomous ride sharing, blader will say it is a L2 ADAS and therefore cry "I was right" until our ears start bleeding.

I think that is what he is counting on. He knows Tesla wants to do the Tesla Network and the odds of them keeping some form of supervision is very high (since Tesla has not demonstrated L4 autonomy) so the odds of a Tesla Network with some form of driver supervision are very good. So odds are very good that he will be able to declare "I was right" even if key details of his theory were wrong. He will still claim that Tesla released the Tesla Network as a L2 ADAS therefore he was right.

As for the ”full autonomy” quote. In Tesla speak it is not that simple. They might strive for full autonomy while still in actual fact releasing a Level 2 ADAS system at this time. Not unlike Elon calling NoA as FSD on highways...

I feel like this where the levels of autonomy are misleading. If Tesla cars in the ride sharing network can come pick you up with no driver inside within a certain radius (which Enhanced Summon has already demonstrated it can do) and the car does successfully drive you across town with no driver disengagements and no hands on wheel nags, only the inside camera to monitor the driver that exists in the Model 3, sure that might still be L2 technically because of the requirement for the driver to monitor, but that is a far cry from what most people would consider L2 ADAS.

I want to emphasize the point that Model 3's have that internal camera that could monitor driver attention and therefore remove all hands on wheel nags. So the Model 3's in the Tesla Network may have zero hands on wheel nags. In that case, wouldn't that actually be L3, not L2, since there is no hands on wheel?

 

[OPRCE]

So Elon might pull off another hat trick. I have always considered the initial launch of AP1 to be dangerous only because Tesla allowed up to 30 mins of no nags and put absolutely no restrictions on it. Now they are learning but the 4 deaths that have happened with AP could have been avoided. But since actual full autonomy isn't here and won't be here for a while, although Elon has been saying its "two years away" since 2015.

Elon will no doubt continue to deceive and say that FSD is here. But since he can't deliver, he will instead will do 'FSD' as a L2 system. I have believed this before but now i know i'm right Looks like i was right. Elon will try to launch "Tesla Network" as L2 ADAS. It will be a uber-like Bird/Limes with pick up spots in downtowns and parking lots. The driver being responsible.

It will be similar to AP but for streets. It will require a driver and require the driver to pay attention to supervise the "FSD" system. Here's why Its brazen, brilliant & dangerous but might actually work.

Its brazen because it relies on untrained drivers who ain't familiar with Teslas to be attentive. Its brilliant because it solves the problem of waiting for L5 (L4 everywhere), increases mind-share, adds revenue, app market-share, but continues myth that Teslas are actually self-driving and are the leader. Dangerous because people will die as result.

It will be a Uber-like Bird/Limes/MoGo. There will be two ride options, miles base and time(hours) based.
There will pickup spots which are also drop off spots around downtown areas and parking lots in downtown around the country.

Miles Based Option:

You go to a pick up spot, use your phone and request a ride to your destination, it will select a drop off spot closest to your destination, it will assign you a car which you get in.
And it will drive with you as the driver and will alert you to be fully prepared to take over at any time that you have to accept. They will also use the rear mirror camera to watch you to make sure you don't fall asleep or do anything crazy (would be crazy if they just relied on the wacky steering sensor). Once you get to the pick-up spot that is closet to your destination, it will park and you get out and go about your day.

Time Base Option:

The car will take you to any destination with you supervising. You get out, do whatever you want to do and then when its time to drop the car back, you get in and have it navigate to the closest pickup/drop off.

This is similar to electric scooters bikes that are based on hours(time) based not miles based, so you are allotted say 6 hours and you can take it home, run errands, etc.

This can actually work and allow Elon to still claim FSD, winning mind-shares as people will continue to believe Teslas are self driving, while also making money. It will technically still be "Tesla Network". But there will be casualties.

It's a decent theory, Blader, which could technically work but I don't think Tesla will go this way because:

1. It would require putting out a large fleet of vehicles which would only slowly pay for themselves with user rentals, compared to getting the whole price upfront as a sale. Tesla can scarcely afford it at this point while M3 sales are good and likely to keep growing strongly.

2. People would have to employed to maintain the fleet, another considerable cost.

3. This whole fleet essentially parked outside on the street would be exposed to hater-type vandalism, etc., which makes for bad advertising.

My own feeling is that the Tesla Network is more likely to launch (quite gradually) as a HW3 & FSD L2 service wherein existing owners log in with the car as available to give lifts within a defined area, whoever is closest then gets dispatched from HQ like an Uber to grab any particular fare, which is paid to Tesla via the user's smartphone and the driver's cut collected and transferred monthly to bank account.

This alternative approach would:

4. neatly circumvent the three hitches mentioned.

5. provide drivers accustomed to the system, thus reducing the probability of casualties from a fresh stream of novice drivers daily.

6. allow Tesla to build up all the necessary infrastructure/backend until L4 arrives and the drivers can be dispensed with, possibly around 2025 at very earliest.

7. provide a failsafe business model in case their L4 is delayed to 2030 -- Tesla can still make good money on TN, perhaps snagging a decent chunk of the AV ride-sharing market mindshare before other OEM competitors can launch their own offerings.

8. allow retrofitted Models S & X to join the scheme.

9. perhaps the biggest benefit -- it would provide a continual public demo of the Tesla technology's capabilities/progress, under supervision of competent safety drivers with own skin in the game, which would be a priceless viral PR bonanza essentially for free.

I hope this actually happens later this year, even if not announced on 22nd!

It's a pivot, but what the hell if it well serves the ultimate goals?

My sh*tl*st just got 2 more items.

Would you mind explaining why you felt compelled to spray this bile which adds precisely nothing to the discussion?

 

[mongo]

So Elon might pull off another hat trick. I have always considered the initial launch of AP1 to be dangerous only because Tesla allowed up to 30 mins of no nags and put absolutely no restrictions on it. Now they are learning but the 4 deaths that have happened with AP could have been avoided. But since actual full autonomy isn't here and won't be here for a while, although Elon has been saying its "two years away" since 2015.

Elon will no doubt continue to deceive and say that FSD is here. But since he can't deliver, he will instead will do 'FSD' as a L2 system. I have believed this before but now i know i'm right Looks like i was right. Elon will try to launch "Tesla Network" as L2 ADAS. It will be a uber-like Bird/Limes with pick up spots in downtowns and parking lots. The driver being responsible.

It will be similar to AP but for streets. It will require a driver and require the driver to pay attention to supervise the "FSD" system. Here's why Its brazen, brilliant & dangerous but might actually work.

Its brazen because it relies on untrained drivers who ain't familiar with Teslas to be attentive. Its brilliant because it solves the problem of waiting for L5 (L4 everywhere), increases mind-share, adds revenue, app market-share, but continues myth that Teslas are actually self-driving and are the leader. Dangerous because people will die as result.

It will be a Uber-like Bird/Limes/MoGo. There will be two ride options, miles base and time(hours) based.
There will pickup spots which are also drop off spots around downtown areas and parking lots in downtown around the country.

Miles Based Option:

You go to a pick up spot, use your phone and request a ride to your destination, it will select a drop off spot closest to your destination, it will assign you a car which you get in.
And it will drive with you as the driver and will alert you to be fully prepared to take over at any time that you have to accept. They will also use the rear mirror camera to watch you to make sure you don't fall asleep or do anything crazy (would be crazy if they just relied on the wacky steering sensor). Once you get to the pick-up spot that is closet to your destination, it will park and you get out and go about your day.

Time Base Option:

The car will take you to any destination with you supervising. You get out, do whatever you want to do and then when its time to drop the car back, you get in and have it navigate to the closest pickup/drop off.

This is similar to electric scooters bikes that are based on hours(time) based not miles based, so you are allotted say 6 hours and you can take it home, run errands, etc.

This can actually work and allow Elon to still claim FSD, winning mind-shares as people will continue to believe Teslas are self driving, while also making money. It will technically still be "Tesla Network". But there will be casualties.

Hate is clouding your mind. There is no way that type of system would work.

1. Like the scooters, the cars would end up left anywhere (if the driver can control, the driver will control)
2. Without true FSD they can't:
A) reposition back on demand
B) move to proper parking spots if abandoned (or snow emergency occurs)
C) enforce traffic laws
3. People without cars have no vehicle insurance, yet this allows them to manually take over driving.
4. Too easily manipulated to make it non-useful (van load of uber drivers take cars to low traffic areas)
5. Problems relying on humans is the reason for FSD (see also Uber crash)
6. Doesn't make full use of vehicle or parking resources (round trips are called out)
7. Users may not even have a driver's license (old, young, city dweller with no need)

 

[Reciprocity]

So Elon might pull off another hat trick. I have always considered the initial launch of AP1 to be dangerous only because Tesla allowed up to 30 mins of no nags and put absolutely no restrictions on it. Now they are learning but the 4 deaths that have happened with AP could have been avoided. But since actual full autonomy isn't here and won't be here for a while, although Elon has been saying its "two years away" since 2015.

Elon will no doubt continue to deceive and say that FSD is here. But since he can't deliver, he will instead will do 'FSD' as a L2 system. I have believed this before but now i know i'm right Looks like i was right. Elon will try to launch "Tesla Network" as L2 ADAS. It will be a uber-like Bird/Limes with pick up spots in downtowns and parking lots. The driver being responsible.

It will be similar to AP but for streets. It will require a driver and require the driver to pay attention to supervise the "FSD" system. Here's why Its brazen, brilliant & dangerous but might actually work.

Its brazen because it relies on untrained drivers who ain't familiar with Teslas to be attentive. Its brilliant because it solves the problem of waiting for L5 (L4 everywhere), increases mind-share, adds revenue, app market-share, but continues myth that Teslas are actually self-driving and are the leader. Dangerous because people will die as result.

It will be a Uber-like Bird/Limes/MoGo. There will be two ride options, miles base and time(hours) based.
There will pickup spots which are also drop off spots around downtown areas and parking lots in downtown around the country.

Miles Based Option:

You go to a pick up spot, use your phone and request a ride to your destination, it will select a drop off spot closest to your destination, it will assign you a car which you get in.
And it will drive with you as the driver and will alert you to be fully prepared to take over at any time that you have to accept. They will also use the rear mirror camera to watch you to make sure you don't fall asleep or do anything crazy (would be crazy if they just relied on the wacky steering sensor). Once you get to the pick-up spot that is closet to your destination, it will park and you get out and go about your day.

Time Base Option:

The car will take you to any destination with you supervising. You get out, do whatever you want to do and then when its time to drop the car back, you get in and have it navigate to the closest pickup/drop off.

This is similar to electric scooters bikes that are based on hours(time) based not miles based, so you are allotted say 6 hours and you can take it home, run errands, etc.

This can actually work and allow Elon to still claim FSD, winning mind-shares as people will continue to believe Teslas are self driving, while also making money. It will technically still be "Tesla Network". But there will be casualties.

I think bladder is the one that's 2 years to late with this idea:

2017 Investor Roundtable: TSLA Market Action

 

[diplomat33]

TL/DR Summary: What the Tesla Network ends up looking like depends entirely on how good Tesla's FSD is. If Tesla has achieved true FSD then Tesla's ultimate goal of driverless Teslas acting as robo-taxis becomes realistic. If Tesla does not achieve any meaningful autonomy then the Tesla Network will look more like Uber or Turo with someone in the driver seat no matter what. The key for Tesla is to offer a version of the Tesla Network as soon as possible in order to capitalize on the potential profit but doing so in a way that is safe and reduces Tesla's liabilities. Using current owners as safety drivers or hiring Tesla employees as safety drivers is the best way to do that.

Now, if you want the details, let me try to break it down for you.

Scenario 1: True L4 autonomy
Under this scenario, Tesla achieves true L4 autonomy and the cars can self-drive with no driver inside.

The implementation of the Tesla Network is obvious and straightforward:
1) Tesla could have a fleet of post-lease Model 3's and have them go out driverless picking up customers
2) Current owners could have their cars go out driverless picking up customers while they are doing other things.

Musk has repeatedly described the Tesla Network this way and the latest lease language also supports this description. The only problem, and it's a big one, is that as far as we know, Tesla has not achieved L4 autonomy. And without L4 autonomy, this scenario completely falls apart. All we have right now is a good L2 ADAS with promises that L4 is just a matter of software that is coming. That's a long way to being able to roll out a fleet of autonomous taxis. And a lot of people have serious reservations that Tesla will ever achieve L4 autonomy on the current hardware. It is also worth noting that even if Tesla did achieve L4 autonomy, it would need serious testing and validation before it could be rolled out in a large fleet of autonomous ride-sharing Teslas. Heck, we see serious companies like Waymo that are already at the stage of accepting passengers in their autonomous taxis, still testing and validating. So the idea that Tesla with just a L2 ADAS, with just a future software update, will suddenly roll out a fleet a L4 autonomous ride-sharing cars, is not realistic.

Scenario 2: "partial" L4 autonomy
Under this scenario, the AP3 computer and the latest software actually does bring Tesla pretty close to L4 autonomy. So Tesla would feel confident that the cars actually could do some ride-sharing with minimal or no input from a safety driver that would still be required just in case. The internal camera would be used to monitor the safety driver's attention. The self-driving would be true hands free, no nags.

So here is a couple ways a Tesla Network could work under scenario 2:
1) Current owners act as a safety driver. This would work but would defeat the whole vision of being able to make money with your car while you are off doing something else. But it would offer current owners a way to make extra money on their free time and would keep risk to a minimal.
2) Tesla employees would be hired to be safety drivers for Tesla's post-lease fleet of FSD Model 3 cars. This would work. It would minimize Tesla's liability while also allowing the system to go through further testing and validation towards L4 autonomy. The option would also keep risk to a minimum for Tesla.
3) Tesla makes customers agree to be safety drivers when they get in the car. The problem with this implementation is that it would be very risky for Tesla. Even if the system really were almost L4, the odds of accidents go up the more cars in the fleet. So there would be tremendous risk. It would only maybe work for the fleet of cars that Tesla owns. It would not work for cars that regular Tesla owners own. There is no way that a regular owner is going to let strangers be safety drivers in their car without them present. I certainly would never go for that. Plus, strangers would not be educated about how to use Tesla's system. It would be prone to all kinds of problems. As I said before, I've seen some folks who can't even figure out how to open the doors on the Model 3. So I think it is highly unlikely that you would trust them to monitor the car during autonomous driving when they have no familiarity with a Tesla at all.

Scenario 3: L2 ADAS.
Under this scenario, AP3 and the future software does not really get close to L4 at all and the system is still squarely a L2 ADAS that requires hands on wheel.

So here is a couple ways a Tesla Network could work under scenario 3:
1) Current owners would be required in the driver's seat at all times. Essentially, this would be an Uber type service where someone can request a ride but there is a human driver in the driver's seat. Autopilot would offer the convenience that some of the driving could be automated so the driver could be more relaxed while taking rides around. This scenario would definitely work now but again, it would kinda defeat the purpose of being able to make money while doing something else. However, some owners who have free time, might like this option.
2) Customers could rent Teslas from the Tesla fleet or from current owners. This would be a Turo type service. Customers would be required to be in the driver's seat at all times and sign a waiver. It would have the advantage that customers could learn about Tesla and enjoy the convenience's of Autopilot. This option would definitely work now since it does not require L4 autonomy. It could be done with the current software even. And it could give Tesla and owners more money.

In conclusion, Tesla talks about scenario 1 (full autonomy) and it is clearly what they are shooting for even though it is NOT doable yet. As we go gown to lower scenarios that require less autonomy, we get versions of the Tesla Network that fall short of what Tesla wants are definitely more doable. I think the April 22 event will shed more light on what we will actually get. If Tesla does demo something that comes super close to L4 autonomy, then scenario 2 will be the most likely, and scenario 1 would be a real possibility further down the line. If Tesla demos something that is still far from L4, then scenario 3 will be the most likely. Ultimately, it all depends on how good Tesla's self-driving is and how autonomous it really is.

 

[Bladerskb]

First of all, it's a theory. I don't know how you can present a theory as a "fact" and say you that you know you are right. You did not present any specific evidence, you just laid out the details of your theory.

Second, I see two ways where you might be wrong:
1) It's too dangerous. There is no way you are going to let untrained folks who don't know Teslas supervise the car's FSD. I've given Model 3 rides to my friends and some of them who have never seen a Tesla before, did not even know how to work the door handle on the Model 3!!! I watched them fiddle with the door handle trying to figure it out! You think an owner or Tesla itself is going to leave these folks responsible for supervising FSD?!?! No way! Frankly, that's not sustainable. As soon as accidents start happening or customers have bad experiences, it would be the death of the Tesla Network.

The only way it could maybe work is if Tesla's autonomy is good enough that very little intervention is actually required. Or it could work if Tesla hires safety drivers for the Tesla Network.

2) Tesla continues to actually claim full autonomy is coming. Tesla actually puts in the Model 3 lease:

"Please note, customers who choose leasing over owning will not have the option to purchase their car at the end of the lease, because with full autonomy coming in the future via an over-the-air software update, we plan to use those vehicles in the Tesla ride-hailing network"
Tesla launches Model 3 leases, will keep cars for autonomous Uber-like service after term

How does this square with your theory? So Tesla will use Model 3's after the lease for the Tesla Network. They are promising "full autonomy" but will require strangers to monitor that full autonomy? That seems very unlikely.

Here's my theory: Tesla will hire Tesla employees to be safety drivers. So the cars will drive around autonomously and pick people up but with a Tesla employee as a safety driver.

We will see on the 22nd. Actually we won't because Elon would most likely punt. He likes to obfuscate things because mystery leads to speculation that leads to hype. Most things around Tesla are myths.

Yes it is dangerous, but so was releasing AP1 without requiring driver check for up to 30 mins at times. It was a death trap waiting to happen. So this isn't a precedent for Elon, he has done this already with this entire legal and engineering team firmly against it. So i won't put it past him. Now there are ways to make it less dangerous but then again this is Elon and he might not do none of this because he will think its too cumbersome, the idea of people dying is collateral damage to him.

Some of the things you can do to add safety to it:

1) Majority of routes only in downtown areas (35 MPH and under).
2) Limit left turn use.
3) Absolutely no unprotected left turns on main roads.
4) Limit highway routes.
5) Require each users to watch interactive video training before first time use and re-occurring after every month.
6) Display voice and visual message disclaimer before each ride.
7) Use driver facing camera to enforce driver attention requirement.

Secondly Tesla has been claiming 'full autonomy' is coming since 2015. Its called PR and Tesla is the worst at it.
Tesla also downgrade FSD from a L5 system to a L2 ADAS early this year.

How does this square with your theory? So Tesla will use Model 3's after the lease for the Tesla Network. They are promising "full autonomy" but will require strangers to monitor that full autonomy? That seems very unlikely.

It seemed very unlikely that Tesla will have people supervise "full self-driving" in 2016 when it was being talked about and Elon saying you could cross country summon in 2018, yet here we are in 2019 and "full self-driving" is being introduced as L2 features.

Here's my theory: Tesla will hire Tesla employees to be safety drivers. So the cars will drive around autonomously and pick people up but with a Tesla employee as a safety driver.

They will barely make any money doing that and might lose money. That would basically be a clone of Uber.

 

[Bladerskb]

Possibly, but I think there is one big miss with this.
Why can't I just get in a Tesla, and drive it into the city where it then parks itself?

90% of the fun of something like Bird/Lime is you can just go anywhere without having to worry too much about where you leave them.

True but Its all about the Hype/Marketing/Branding. So it will just be a L2 ADAS but be marketed as "full self-drivinf" and the hyping, marketing and branding of Elon/Tesla would make it seem "revolutionary".

I think bladder is the one that's 2 years to late with this idea:

2017 Investor Roundtable: TSLA Market Action

Its interesting that you got disagreed on that post. shows you that back then saying FSD was L2 would face stiff resistance because people expect FSD to launch in 2018 with the Tesla Network.

I want to emphasize the point that Model 3's have that internal camera that could monitor driver attention and therefore remove all hands on wheel nags. So the Model 3's in the Tesla Network may have zero hands on wheel nags. In that case, wouldn't that actually be L3, not L2, since there is no hands on wheel?

Its always funny to watch you flip flop when it suites you.

 

[diplomat33]

True but Its all about the Hype/Marketing/Branding. So it will just be a L2 ADAS but be marketed as "full self-drivinf" and the hyping, marketing and branding of Elon/Tesla would make it seem "revolutionary".



Its interesting that you got disagreed on that post. shows you that back then saying FSD was L2 would face stiff resistance because people expect FSD to launch in 2018 with the Tesla Network.



Its always funny to watch you flip flop when it suites you.

I am not flip flopping. I am asking the question: would that scenario of no nags be L3 or L2?

But you missed my earlier point. You keep saying that it will be L2 ADAS but Musk will falsely hype it up as revolutionary FSD. You might be right that it would technically be L2 ADAS but if it is L2 ADAS that actually drives people around with no nags and no hands on wheel that would be revolutionary indeed. That would not be hype. It would be quite an achievement.

 

[Bladerskb]

1. It would require putting out a large fleet of vehicles which would only slowly pay for themselves with user rentals, compared to getting the whole price upfront as a sale. Tesla can scarcely afford it at this point while M3 sales are good and likely to keep growing strongly.

Which is why they will use "used leased model 3s" to do it. while they are waiting to be sold/re-leased.

2. People would have to employed to maintain the fleet, another considerable cost.

But you can have 1 person for every 10 car for example, so you only need to pay one person rather than 10 people.

3. This whole fleet essentially parked outside on the street would be exposed to hater-type vandalism, etc., which makes for bad advertising.

Which is what sentinel is there for, plus (reserved and marked) street parking isn't the only place it will be, it could also be in first floor of parking garages, reserved and marked parking lots of stores.

My own feeling is that the Tesla Network is more likely to launch (quite gradually) as a HW3 & FSD L2 service wherein existing owners log in with the car as available to give lifts within a defined area, whoever is closest then gets dispatched from HQ like an Uber to grab any particular fare, which is paid to Tesla via the user's smartphone and the driver's cut collected and transferred monthly to bank account.

That would literally be a Uber clone. I don't think people bought a $45k+ car in cash to be a uber driver. These people would already have money and would be they already have jobs.

This alternative approach would:

6. allow Tesla to build up all the necessary infrastructure/backend until L4 arrives and the drivers can be dispensed with, possibly around 2025 at very earliest.

7. provide a failsafe business model in case their L4 is delayed to 2030 -- Tesla can still make good money on TN, perhaps snagging a decent chunk of the AV ride-sharing market mindshare before other OEM competitors can launch their own offerings.

9. perhaps the biggest benefit -- it would provide a continual public demo of the Tesla technology's capabilities/progress, under supervision of competent safety drivers with own skin in the game, which would be a priceless viral PR bonanza essentially for free.

It's a pivot, but what the hell if it well serves the ultimate goals?

It is absolutely a pivot, whatever way it shakes out. But most adoring fans and shareholders won't see it as that.
Frankly I believe the rider as the driver supervisor as being the only path that provides incentives to riders as the cost would be lower than getting an Uber/Lyft (with less convenience ofcourse), but also provides the most profits because you don't have to pay drivers/employees. They can undercut uber/lyft drastically.

Think about it, majority of uber/lyft rides are around the downtown areas. Downtown is hectic for parking. People avoid downtown because of parking. Even I would take a car that can get me close to my destination where all i had to do was to walk the rest of the way. Especially if its way cheaper than uber/lyft.