The Use by Other EVs of Tesla Supercharges?

[MDMGSO47]

I know this must have been discussed before, but I did a search and could not find the answer.

What, if anything, prevents other EVs from using Tesla Superchargers? I assume an adapter is necessary to allow the Tesla supercharger connection to fit another vehicle's charge port? Even if there are adapters available, couldn't Tesla prevent the use by the supercharger recognizing that it has been connected to a vehicle of another brand?

 

[TexasEV]

You answered your own question.

 

[deonb]

Same thing that prevents Model S 60s that didn't pay for Supercharging from charging at them.

 

[Maximus8]

...couldn't Tesla prevent the use by the supercharger recognizing that it has been connected to a vehicle of another brand?

Yup and yessir

 

[Saghost]

I know this must have been discussed before, but I did a search and could not find the answer.

What, if anything, prevents other EVs from using Tesla Superchargers? I assume an adapter is necessary to allow the Tesla supercharger connection to fit another vehicle's charge port? Even if there are adapters available, couldn't Tesla prevent the use by the supercharger recognizing that it has been connected to a vehicle of another brand?

Well, first the other car would have to speak Supercharger. Communication is initiated with a standard J1772 "go digital" 5% duty cycle on the pilot - but then the car reaches out to the SpC with a unique Tesla communication sequence.

If they figure that out, they still aren't home free - one of the few things that TMC members have teased out on the SpC-car link is that the first thing the car passes to the SpC is the VIN - presumably to check against a whitelist on a Tesla server.

 

[AB4EJ]

Well, also, the SC connector is unique to Tesla. I have never heard of anyone making an adapter that would let you plug a SC cable into a different type of vehicle. Probably nobody makes these because they wouldn't work anyway due to the communication requirements.

 

[SomeJoe7777]

Someone made an adapter that uses a Tesla-compatible socket and a J1772 plug. He has used it to charge a BMW i3 from a HPWC:

BMW I3 Charging on Tesla HPWC

This works because the HPWC protocol is the same as J1772. But it will not work on a Supercharger because of the Tesla-specific communication.

While there has been a lot of speculation about how Superchargers can prevent unauthorized vehicles from charging, I don't think the Superchargers are on a live network and authenticate the car via a central server. Many Superchargers are in the middle of nowhere and live communication to a central server would be difficult or impossible.

Instead, I speculate there are two separate authentication mechanisms:

1. The car does not allow supercharging unless that feature has been enabled in the car. This is how the 40s and 60s that aren't supercharger-enabled work. They simply won't tell a supercharger to begin charging.
2. I'm betting that the communication packet that the car sends to the supercharger contains:

a. The car's VIN
b. Date and time (to prevent replay attacks)
c. Whether the car is authorized to supercharge or not.
d. Maximum power the car can take.
e. Amount of power the car is requesting.
f. This entire packet is digitally signed with a private key embedded in the car's firmware, and possibly with a car-specific key.

The supercharger would then validate the digital signature using the other half of the signing key.

So even if you make an adapter and reverse-engineer the proper communication protocol, there is no way for you to sign this packet without the private key, so you'll never get the supercharger to supply power to you.

This would also allow Tesla to revoke permission to supercharge from another manufacturer who bought into the supercharger network if that manufacturer's car firmware was compromised and the private key leaked. You just revoke that key from the supercharger firmware.

 

[Saghost]

S
While there has been a lot of speculation about how Superchargers can prevent unauthorized vehicles from charging, I don't think the Superchargers are on a live network and authenticate the car via a central server. Many Superchargers are in the middle of nowhere and live communication to a central server would be difficult or impossible.

I certainly don't know that Tesla is authorizing each session through their servers - but I do know that many (all?) supercharger sites communicate with Tesla's servers in real time - as the display they've had in the Fremont factory for a few years now shows (see first picture in the article, with real time usage.) Elon promised to provide the data in car at some point in a firmware update:

Tesla will finally release real-time Supercharger availability data

 

[badbot]

If someone hacks the supercharger connection and enables charging to his car then 120A 400V rushes down the cable and his batteries start smoking teaching him karma. that and 10 grand get him back on the road with new parts.

 

[Bangor Bob]

Same thing that prevents Model S 60s that didn't pay for Supercharging from charging at them.

A software flag in the car? At the moment it's quite likely that the superchargers are trusting the car to be "truthful."

 

[Bangor Bob]

2. I'm betting that the communication packet that the car sends to the supercharger contains:

a. The car's VIN
b. Date and time (to prevent replay attacks)
c. Whether the car is authorized to supercharge or not.
d. Maximum power the car can take.
e. Amount of power the car is requesting.
f. This entire packet is digitally signed with a private key embedded in the car's firmware, and possibly with a car-specific key.

The supercharger would then validate the digital signature using the other half of the signing key.

Logs have been taken of supercharger sessions. They're not (quite) that complicated, and not encrypted.

 

[deonb]

A software flag in the car? At the moment it's quite likely that the superchargers are trusting the car to be "truthful."

It's certainly possible it's just in the clear, but I imagine it's 80% likely to be signed.

 

[SomeJoe7777]

Logs have been taken of supercharger sessions. They're not (quite) that complicated, and not encrypted.

Would love to see some of those logs. And true, they may not be encrypted, as that's not terribly necessary, but I would say there has to be some kind of authentication more than just the car's decision.

I certainly don't know that Tesla is authorizing each session through their servers - but I do know that many (all?) supercharger sites communicate with Tesla's servers in real time - as the display they've had in the Fremont factory for a few years now shows (see first picture in the article, with real time usage.) Elon promised to provide the data in car at some point in a firmware update.

Will that be for all superchargers, though? Many of them that are within a major city can easily communicate over wireless just like the car does, but at others in desolate locations I don't know that the wireless signal from the carrier is receivable.

Furthermore, even if there is real-time communication, you can't consider it 100% reliable, so you can't be authorizing cars through it to supercharge. You don't want to have a supercharger unable to charge cars just because you can't communicate with HQ, so any authentication the supercharger is doing has to be local.

I suppose if each supercharger has communication capability, but it's not guaranteed to be up 100% of the time, then you could periodically (once a day) download authorized VINs to each supercharger, and then the supercharger can simply authenticate each car using the VIN. This would really open the chargers up to hacking though, since it would be trivial to spoof the supercharger protocol using a known good VIN.

 

[NewCow]

I'm not an expert, but my understanding is that most other EVs also can't technically handle the charge. Most cars top out with DC fast charging around 50 kW, while Supercharging is 120kW.

 

[deonb]

I'm not an expert, but my understanding is that most other EVs also can't technically handle the charge. Most cars top out with DC fast charging around 50 kW, while Supercharging is 120kW.

Supercharging is max of 120kW, but the car can draw as much as it can handle from it. A Model S at the top end of the taper curve draws far far less than 50kW from a Supercharger.

 

[ElectricTundra]

Authentication can be local to each supercharger with occasional updates from a central host of authorized VINs, etc. This eliminates most network outage problems as the local host always operates with whatever its latest data is.

Since these are fixed locations I'd think they'd be able to either beam their signal towards two or three cell towers or use satellite if they're in the middle of nowhere. A bit of delay from satellite shouldn't be problematic for this application.

 

[Saghost]

Would love to see some of those logs. And true, they may not be encrypted, as that's not terribly necessary, but I would say there has to be some kind of authentication more than just the car's decision.



Will that be for all superchargers, though? Many of them that are within a major city can easily communicate over wireless just like the car does, but at others in desolate locations I don't know that the wireless signal from the carrier is receivable.

Furthermore, even if there is real-time communication, you can't consider it 100% reliable, so you can't be authorizing cars through it to supercharge. You don't want to have a supercharger unable to charge cars just because you can't communicate with HQ, so any authentication the supercharger is doing has to be local.

I suppose if each supercharger has communication capability, but it's not guaranteed to be up 100% of the time, then you could periodically (once a day) download authorized VINs to each supercharger, and then the supercharger can simply authenticate each car using the VIN. This would really open the chargers up to hacking though, since it would be trivial to spoof the supercharger protocol using a known good VIN.

If spoofing became an issue, Tesla has lots of easy options - remember, they appear to have near real time communication with both sides of the charge - if anomalies started cropping up, they could query the car with that VIN and see what it was up to.

It'd be hard to justify in court why you were spoofing a Tesla VIN...

 

[ElectricTundra]

I'd think Tesla could fairly easily use a bit of EMV protocol for auth. Secure and has provisions to continue operating without network.

Charging is also something that wouldn't be too big of a deal if a few unauthorized people are allowed to charge when the network is down so worst case the chargers could simply default to 'allow charge' if they don't get a response from the local host or central host within some specified period of time.

 

[Topher]

Many Superchargers are in the middle of nowhere and live communication to a central server would be difficult or impossible.

Sorry, but what century do you think you are living in?

But really, there are about 100,000 authorized VINs at 4 bytes per, so 1/2 a meg... Let me know when we get up to the smallest SD card you can buy.

Thank you kindly.

 

[electracity]

Sorry, but what century do you think you are living in?

But really, there are about 100,000 authorized VINs at 4 bytes per, so 1/2 a meg... Let me know when we get up to the smallest SD card you can buy.

Thank you kindly.

Exactly. Musk came from paypal. The car and the supercharger network has likely had a secure handshake since day 1.

I would guess that the Toyota seen supercharging is a Tesla owned vehicle running the Model 3 drivetrain or other test equipment.