Sentry mode

[Wol747]

I had occasion to park my MS at a relative's in Manly for 3 nights and left sentry mode on. We all know it's a battery guzzler but I was quite shocked when it used 48Km worth of range per 24 hrs. It WAS parked where we walked past it probably 9 or 10 times in the three days but.......

 

[Vostok]

Yep, Sentry is an absolute power hog, so I only turn it on when I am nervous about where my Model 3 is parked. Typical power usage is between 1 and 2 km per hour.

The most interesting things I have caught on Sentry are people checking out the car. But most people don’t take any particular notice of it.

 

[RareEarth]

Always have it on when parked away from home.
Although watching sentry mode footage from a busy carpark can be a bit cringeworthy - car doors opening, people walking between cars swinging handbags, shopping bags etc. I always try to park away from the busiest areas but sometimes not possible.
So many events can be generated from just a short stay in a busy carpark, not possible to watch it all.
The in-car viewer could do with some improvements though such as speeding up the replay, ability to bulk delete files (apart from just reformatting the drive).

 

[Wol747]

It could also do with a better selector button - we are usually only interested in the bit around the red dot but my finger covers everything and makes it impossible to select that part accurately. It certainly needs the ability to make selections and delete them, as you say.
I'm not sure whether triggering a recording actually uses more electricity - I think the computer must be on all the time in Sentry mode even if the screen's black.
It's all a bit of a fudge from the usability point of view.

 

[Maximillan]

It could also do with a better selector button - we are usually only interested in the bit around the red dot but my finger covers everything and makes it impossible to select that part accurately.

I'm still puzzled as to why I have to watch 1.5 minutes of nothing before I see the event that triggered the sentry. I'm quietly hoping someone here will point out that I'm doing it wrong.

It's particularly difficult when some other car has their indicator on and waiting for you to pull out of the parking spot. Surely Tesla could simply give us the 5 secs before the red dot, and then upon viewing said video provide two options:
1. Save [and next video]
2. Delete [and next video]

 

[atj777]

I'm still puzzled as to why I have to watch 1.5 minutes of nothing before I see the event that triggered the sentry. I'm quietly hoping someone here will point out that I'm doing it wrong.

It's particularly difficult when some other car has their indicator on and waiting for you to pull out of the parking spot. Surely Tesla could simply give us the 5 secs before the red dot, and then upon viewing said video provide two options:
1. Save [and next video]
2. Delete [and next video]

On my Model 3, it typically starts playing around 10 seconds before the red dot. It never plays the clip from the beginning.

 

[Vostok]

On my Model 3, it typically starts playing around 10 seconds before the red dot. It never plays the clip from the beginning.

Same. One can also tap the timeline bar at a different spot to advance or retreat the playhead.

 

[louco73]

I did an experiment with Sentry mode last week and thought I'd share the results here.

I flew to Sydney for a business trip on Sunday the 20th and returned on Friday the 25th. I checked in on the car each day and took a screen shot to record the state of charge. Here is the data with the assumption the battery is 79.5 kWh as per EVSpecifications:

Feb 20, 10:32, 61%, 48.5 kWh
Feb 21, 13:44, 56%, 44.5 kWh , ~147 W consumption
Feb 22, 07:44, 53%, 42.1 kWh, ~133 W consumption
Feb 23, 10:26, 46%, 36.6 kWh, ~206 W consumption
Feb 24, 11:24, 39%, 31 kWh, ~224 W consumption
Feb 25, 13:38, 30%, 23.9 kWh, ~273 W consumption
Feb 25, 19:44, 29%, 23.1 kWh, ~133 W consumption

There were 50 Sentry mode events during the entire period. I've not looked at them, so not sure of the activity spread across the days. I also realized that I had the cabin heat protection mode with AC set to on, so that might have fired up at some point.

My car is a 2020 Model 3, Performance. It would be interesting to see if the new 2022 cars are more efficient?

On normal days that I travel from the Gold Coast to work in Brisbane I would use about 20% each way. Unfortunately that calculation went out the window when we arrived in Brisbane and it was raining heavily. Adding my wife and luggage to the equation was not helping either. Thus, soon after the Gateway Bridge the car was complaining and asking me to drive below 90 km/h. Then after the M3/M1 merge the car was saying we would not make it home and to drop speed further. I changed the destination to Helensvale Railway Station, thinking I could get a quick top up there and kept going at 90 km/h. With the rain starting to ease as we headed south I tried dropping the speed to 85 km/h and set the destination to home. The car said we would arrive with 3%.

I stuck to the left lane and kept the car set to 85 km/h for the rest of the journey. Surprisingly I still had to pass a couple of cars that were going slower, presumably because of the weather. I kept asking my wife to check the navigation status and it dropped to 2% and then 1% during the trip. I turned off the A/C once we got on Smith St. and accelerated very slowly from any lights. Thankfully we made it home without needing a charge and 1% left.

Even at those low speeds the consumption was 205 Wh/km with a normal trip to Brisbane and back using ~165 Wh/km.

Lesson - Charge to 100% before going to the airport if you want to use Sentry Mode .

 

[Wol747]

>>I stuck to the left lane and kept the car set to 85 km/h for the rest of the journey. Surprisingly I still had to pass a couple of cars that were going slower, presumably because of the weather. I kept asking my wife to check the navigation status and it dropped to 2% and then 1% during the trip. I turned off the A/C once we got on Smith St. and accelerated very slowly from any lights. Thankfully we made it home without needing a charge and 1% left.<<

I’m just having a soothing cuppa while my bum unclenches…….

 

[NoTimeToLose]

Is it correct that in Australia, even with premium connectivity the Sentry Mode won’t send live video to the Tesla app?

If so, any idea why - and is this likely to change any time soon?

 

[Maximillan]

That is correct.

There has been plenty of speculation as to why, nothing has been confirmed with regards to why or when it'll change..

 

[moa999]

any idea why - and is this likely to change any time soon

Like many things it's US first on this feature (and now I think Canada).

Some things move to international markets quickly, others can take years (eg FSD beta)

 

[NoTimeToLose]

Looks like buying a Tesla means I'll just have to work at being more patient. Bonus self-help opportunity!

Thanks for the info, folks.

 

[Vostok]

Lesson - Charge to 100% before going to the airport if you want to use Sentry Mode .

Sentry is an absolute power hog and has always been thus. So yes, if you plan to have sentry mode on when parking somewhere for multiple days, charge the car to 90% or 100% before leaving.

If you had cabin overheat protection on that would use even more power during the day. I have never used COP, so I don’t need to remember to turn it off.

A parked Model 3 that has COP and Sentry off, is not plugged in, and is not woken up by anxious owners checking the App every day, will only lose about 1 km of range per day. So it could be parked for more than a year.

 

[RedSky1]

Sentry has always been a handy side benefit of the self driving tech, rather than a ground up security solution. Due to that compromise they need to run a mini supercomputer constantly that consumes 200-250 watts constantly while on.

Ideally it would have separate cameras or a split feed from those cameras that run into the MCU however the Atom MCU may not have had enough processing power to take 4 camera feeds, monitor the feeds for changes and compress it in real time and then also continue to run the other car MCU functions.

I would guess the Ryzen MCU may have the processing power to do the job and in doing so would use less than 100 watts to do the same thing, but at this point it would just be easier to continue to keep the existing status quo and self driving systems almost identical across cars.