Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tesla app unable to wake car

This site may earn commission on affiliate links.
It's unreasonable to own a vehicle with constant software updates and not to expect occasional 'bugs'. I've been an Apple user for computers and Adobe user for photography editing for many years. It's just a fact of life that occasionally we get bugs. I often don't upgrade software to new versions until I've checked forums to make sure the software is pretty well bug free.
I agree. If I gave the impression that I never expected bugs it was unintended. My post was to inform the community that Tesla are aware of the problem and working on a fix.
 
Further update from Tesla:

Screenshot_20230530_120840_Tesla.jpg
 
Well, how else could you communicate with a car that is not otherwise network attached when it is asleep? Using an SMS is the technically simplest and most elegant solution.
I disagree - i mean it is network attached - because it's listening for an SMS.. i doubt there is any real world difference on a device with a 84kwh battery whether it's listening on GSM or 4G - it's entirely possible that it's a more universal solution for a worldwide car though - could use pager tech if you really wanted low consumption.

Does make me think that it wouldn't work where there's no GSM but is wifi.. (just thinking out aloud)
 
i mean it is network attached - because it's listening for an SMS.
The difference is that the way they do it they don’t need to have the MCU running. That is they can receive a SMS by having only just barely enough compute power active (then if it is an appropriate Wake Up one then it un-sleeps the MCU which then brings up the rest of the car).

If the MCU is running - to actually maintain a data connection or do any other interaction with the mothership to wait to be woken up then the car isn’t “asleep” (ie it will be just like when sentry mode is on as far as hourly power draw).

They need to have the MCU not running and they need to still be able to wake the car remotely.
Does make me think that it wouldn't work where there's no GSM but is wifi..
That is correct. If you have your car connected to wifi and let it sleep (ie sentry mode off, not charging, don’t touch the app or anything else that will keep it awake for 15 minutes or so) you will see - if your access point supports showing you the connected client devices - that the car disconnects from the wifi network when it is asleep. The remote Wake Up can only be sent via SMS (not wifi).
 
  • Like
Reactions: Vostok
The difference is that the way they do it they don’t need to have the MCU running. That is they can receive a SMS by having only just barely enough compute power active (then if it is an appropriate Wake Up one then it un-sleeps the MCU which then brings up the rest of the car).

If the MCU is running - to actually maintain a data connection or do any other interaction with the mothership to wait to be woken up then the car isn’t “asleep” (ie it will be just like when sentry mode is on as far as hourly power draw).
I'd assume the data would be a separate board - it probably does require one of the control computers to send/receive meaningful data - but thats probably by design more than good practice - to be honest it would be better for it to be some sort of management/diag board - like an ILO/IDRAC port something that could do meaningful remote diag without the control ECU being fully functional. I was pondering earlier - i suspect it might be a trade off with shedding load/responsibility for the wake up - it's relatively easy to get a cloud application to flick off an SMS - it's probably a bit more complicated to spawn off a process find the car and deliver a packet to get it's attention or flag it for wake up when it checks in - vs dropping 10c to the carrier and having them manage the delivery - delayed or immediate.
 
I'd assume the data would be a separate board - it probably does require one of the control computers to send/receive meaningful data - but thats probably by design more than good practice - to be honest it would be better for it to be some sort of management/diag board - like an ILO/IDRAC port something that could do meaningful remote diag without the control ECU being fully functional.
It comes down to: the cellular modem can understand the 3G control channel messages, but it doesn't do higher level networking like TCP/IP. They could add a separate very low-power board that can do TCP/IP, but Tesla are all about minimising the part count.
 
How many other Australian customers would have 60,000+ SIM cards in the field on the one account?

The four major banks each have hundreds of thousands of SIM cards in the field. You know those battery powered EFTPOS terminals that just about every shop has? Every single one of them has a SIM card in it.

Then there’s utilities such as power and water with smart meters that use the cellular network - hundreds of thousands more there (not all of them use the cellular network - some use LoraWAN).

I believe Telstra has around 4 million active IoT/M2M services. 60,000 services barely touches the sides.
 
I disagree - i mean it is network attached - because it's listening for an SMS

No, a cellular device in idle mode does not have an IP address and therefore it is not “network attached”. If it was network attached, it would not need to be paged, but instead could be directly communicated with over TCP/IP.

Cellular devices in idle mode need to be powered up to listen for incoming pages, but that does not take much power.
 
  • Informative
Reactions: Sir Surfalot
How many other Australian customers would have 60,000+ SIM cards in the field on the one account?

$ value they may not be as large as some corporates as I assume IoT pricing is somewhat cheaper than voice + data packages.
Transport agencies? (Variable message signs, school zones, smart card validators on vehicles, GPS / Location trackers on vehicles)

Financial institutions? (EFTPOS terminals, ATMs)