Possible hidden costs

[ModelNforNerd]

Although, I can imagine a world where they continue uploading data via wifi and mobile networks but use satellite for downloading updates. This would be trivial and is how satellite TV and Radio software updates come in. There's no need to maintain constant connection and if you are getting gigabit rates then you'd only require short bursts for new software updates, etc.

It all depends on how much the current data is costing them whether or not it's cost effective to implement a satellite solution (would need antennas on every car).

but you could kiss Spotify and the hopes of sharing constant real-time data with other connected vehicles around you goodbye.

 

[JeffK]

but you could kiss Spotify and the hopes of sharing constant real-time data with other connected vehicles around you goodbye.

Depends... they can pick and choose which data goes through which connection, that's trivial. I'm talking about large updates such as cached maps or software updates.

as far as spotify, it'd work just as good as XM/Sirius radio works now... and SpaceX projects latencies between 25 and 35ms. (You'd still be limited by mobile uplink though)

There'd also be so many satellites that it's possible you can have clear view of one even if surrounded by buildings... (4,425 which is insane)

heck, since it's 24 GHz phased array antenna you could mount one on the roof of the car and get two way communications if you really wanted to.

 

[ModelNforNerd]

Depends... they can pick and choose which data goes through which connection, that's trivial. I'm talking about large updates such as cached maps or software updates.

as far as spotify, it'd work just as good as XM/Sirius radio works now... and SpaceX projects latencies between 25 and 35ms. (You'd still be limited by mobile uplink though)

There'd also be so many satellites that it's possible you can have clear view of one even if surrounded by buildings... (4,425 which is insane)

heck, since it's 24 GHz phased array antenna you could mount one on the roof of the car and get two way communications if you really wanted to.

Let's get the car into our driveways before we start worrying about how we're going to support it from orbit....

 

[JeffK]

Let's get the car into our driveways before we start worrying about how we're going to support it from orbit....

It'll probably be in version 2+... don't want too much hubris.

 

[ModelNforNerd]

It'll probably be in version 2+... don't want too much hubris.

exactly. people were rather upset when "falcon wing doors" caused Model X delays.

can you imagine the outcry for "insufficient satellite constellation"?

 

[Garlan Garner]

Sorry you believe that I am spreading rumors. Maybe I wasn't clear in my original statement. The agreement with AT&T probably means that Tesla either pays them a fee for every car sold OR they pay a rate to AT&T up to X amount of cars sold and then ATT can renegotiate the rates with Tesla.
Either way I am ASSUMING that is will not be FREE for Tesla to just add all of these M3s to AT&Ts network. Either Tesla will need to eat the cost or we the consumer will end up paying for it.
What I was trying to state above is that ONE of the ways Tesla is likely to assist in reaching their low price point is to charge the buyer for data for use of streaming radio and things of the sort.

I currently have Viper SmartStart for 2 of my ICE cars. I can remotely start my either of my cars...lock the doors...open the trunk via my phone - no matter where they are. My wife locked the keys in her car here in Chicago and I was able to unlock the car from Florida. Viper charges me $21 per car for 3 years via AT&T. Its an extremely low signaling/data requirement and places miniscule to 0% load on their network on a monthly basis.
If Tesla baked in $1K in each car for such a feature....then we would have paid for that feature for a lifetime.

 

[dsvick]

I currently have Viper SmartStart for 2 of my ICE cars. I can remotely start my either of my cars...lock the doors...open the trunk via my phone - no matter where they are. My wife locked the keys in her car here in Chicago and I was able to unlock the car from Florida. Viper charges me $21 per car for 3 years via AT&T. Its an extremely low signaling/data requirement and places miniscule to 0% load on their network on a monthly basis.
If Tesla baked in $1K in each car for such a feature....then we would have paid for that feature for a lifetime.

The connectivity in a Tesla is used for significantly more than unlocking doors and starting it though. There is all of the driving data being passed back and forth, software updates, streaming media, map updates, and other internet features that you can do. Granted, some of it is probably done via wifi whenever possible, but not all of it.

 

[Garlan Garner]

The connectivity in a Tesla is used for significantly more than unlocking doors and starting it though. There is all of the driving data being passed back and forth, software updates, streaming media, map updates, and other internet features that you can do. Granted, some of it is probably done via wifi whenever possible, but not all of it.

My point is this: The data required for a car is minimal...especially when WIFI is a requirement for updates and such. The data used to transfer information back and forth between a car can actually be done via SNMP traps ( machine to machine texts ). The required bandwidth for that is next to nothing.

 

[ModelNforNerd]

My point is this: The data required for a car is minimal...especially when WIFI is a requirement for updates and such. The data used to transfer information back and forth between a car can actually be done via SNMP traps ( machine to machine texts ). The required bandwidth for that is next to nothing.

What we don't know yet is how much data APv2.0 will be sending out. Remember, even if you don't purchase the option, it's always running in "learning mode" and sharing its data.

I'm going to assume that with the additional sensors, there's a significant amount of data that will need to be moved.

 

[Garlan Garner]

What we don't know yet is how much data APv2.0 will be sending out. Remember, even if you don't purchase the option, it's always running in "learning mode" and sharing its data.

I'm going to assume that with the additional sensors, there's a significant amount of data that will need to be moved.

"WE" don't need to know how much data AP2.0 is and or will be sending out.

AP2.0 data is not what any Tesla owner wants to pay for (maybe some rich Tesla fans might want to...or have been forced to ).... however If Tesla wants 2.0 data...from the M3 platform they should to pay to get it.

Maybe this is part of ELONS statement - Thanks MS and MX owners for paying for the M3. (thanks for paying for us to get all of that data so far. )

 

[ModelNforNerd]

"WE" don't need to know how much data AP2.0 is and or will be sending out.

As long as the AP data is metered separately (which I imagine it would be...), then we don't need to know, especially right now, since Tesla owners aren't paying for data anyway.

If the data IS currently baked into the price of the car, I wonder if Tesla will "decouple" it to get to the $35,000 base price. I would not be shocked in the least if they did.

GM and Audi already make you pay for your in-car 4G data after an initial 6 months for free.

Hmmmm....

 

[dsvick]

My point is this: The data required for a car is minimal...especially when WIFI is a requirement for updates and such. The data used to transfer information back and forth between a car can actually be done via SNMP traps ( machine to machine texts ). The required bandwidth for that is next to nothing.

Streaming music, internet use, and AP sensor data will not be minimal. While SNMP traps may play a role in sending notifications to Tesla about the car, they are not intended for two way communication, even regular SNMP communication is mostly for small messages. While it may be using that for things like unlocking doors, remote starting, and turning on the AC I doubt it is being used for much more than that.

 

[Garlan Garner]

As long as the AP data is metered separately (which I imagine it would be...), then we don't need to know, especially right now, since Tesla owners aren't paying for data anyway.

If the data IS currently baked into the price of the car, I wonder if Tesla will "decouple" it to get to the $35,000 base price. I would not be shocked in the least if they did.

GM and Audi already make you pay for your in-car 4G data after an initial 6 months for free.

Hmmmm....

My question is this....What function of a Tesla car....is for the owner? What function of a Tesla car is for Tesla?

Updates? That's Teslas obligation
Fixes? That's Teslas obligation
AP2.0 data? That's Teslas obligation

Remotely heating the cabin - Owner ( if they choose)
Remotely ......? Owner ( if they choose ) .

Add to the list as you may...and lets see how much Tesla is obligated to and how much the owner is responsible for.

 

[dsvick]

As long as the AP data is metered separately (which I imagine it would be...), then we don't need to know, especially right now, since Tesla owners aren't paying for data anyway.

If the data IS currently baked into the price of the car, I wonder if Tesla will "decouple" it to get to the $35,000 base price. I would not be shocked in the least if they did.

GM and Audi already make you pay for your in-car 4G data after an initial 6 months for free.

Hmmmm....

Depending on what agreement Tesla has with the data provider, yes decoupling maybe their best bet, in which case, the owner would be responsible for their music and internet surfing.

 

[Garlan Garner]

Streaming music, internet use, and AP sensor data will not be minimal. While SNMP traps may play a role in sending notifications to Tesla about the car, they are not intended for two way communication, even regular SNMP communication is mostly for small messages. While it may be using that for things like unlocking doors, remote starting, and turning on the AC I doubt it is being used for much more than that.

Oh NO.....SNMP traps are loaded in where I work. They are large packeted data transmissions. OID's are being added daily. Up to 1 million per second from most of our switches.

That statement you made about SNMP traps is certainly NOT true.

Now some of the big IFs are -
"IF" Tesla has negotiated unlimited data per car with AT&T, then the amount of data being used per user is negligible.
"IF" Tesla has limited the unlimited data per car to something amazingly slow, then the amount of data being used per car might be negligible.
"IF" Tesla has set up its data to ONLY be transferred via WIFI, then the amount of data being used per car might be negligible.
"IF" any single or combination of anything mentioned above are in play, then the amount of data being used per car might be negligible.
"IF" <-------- insert your own IFs here.

 

[Garlan Garner]

Depending on what agreement Tesla has with the data provider, yes decoupling maybe their best bet, in which case, the owner would be responsible for their music and internet surfing.

I'll put it this way.

M3 owners are not the folks Tesla wants to start charging data for. That's what our current govt. in the US are doing concerning taxes on a mass scale.

Who's paying for it now? BAU in my opinion. Tesla isn't complaining.

 

[ModelNforNerd]

I'll put it this way.

M3 owners are not the folks Tesla wants to start charging data for. That's what our current govt. in the US are doing concerning taxes on a mass scale.

Who's paying for it now? BAU in my opinion. Tesla isn't complaining.

some of it may also depend on the data carrier's infrastructure. Is AT+T 4G/LTE robust enough to have triple the amount of Teslas, each beaming exponentially more data than cars currently on their network?

 

[Garlan Garner]

some of it may also depend on the data carrier's infrastructure. Is AT+T 4G/LTE robust enough to have triple the amount of Teslas, each beaming exponentially more data than cars currently on their network?

I understand what you are saying however I can more than assure you that capacity will never be a problem with major carriers. I won't delve into why - because I work for one of them.

 

[ModelNforNerd]

I understand what you are saying however I can more than assure you that capacity will never be a problem with major carriers. I won't delve into why - because I work for one of them.

overall capacity may not be. but there are occasions where localized saturation becomes an issue. It's why they have to wheel in mobile arrays or set up other temporary backhauls for large events.

I imagine sections of California might become a real PITA to use an AT+T phone soon, between the Audis, GMs, Teslas, and cell phones in small areas all pulling data at the same time.

 

[Garlan Garner]

overall capacity may not be. but there are occasions where localized saturation becomes an issue. It's why they have to wheel in mobile arrays or set up other temporary backhauls for large events.

I imagine sections of California might become a real PITA to use an AT+T phone soon, between the Audis, GMs, Teslas, and cell phones in small areas all pulling data at the same time.

Local saturation has nothing to do with Tesla. Even if it is an issue.

I want to keep this on point.