Sentry Mode on Model 3 vs Model X

[Need]

I had Sentry Mode on my Model 3 (5.15) for a while and had been using it for weeks. On my Model X, I just got the 8.3 update and tested it yesterday. I noticed there is a difference between the 2 cars. With Model 3, even if I have the phone (as key) with me, Sentry Mode will still trigger if I come near the car. As a result, it will always trigger and record an event when I got back to car. I found that out early on and I have to remotely turn off Sentry Mode on the app before returning to car to avoid an unnecessary recording.

On the Model X, if I have the keyfob with me, it does not trigger the recording at all. I am guessing it is a side effect of the keyfob unlocking the car before I could get near the Sentry Mode detection zone. Or maybe this is an improvement on software 8.3 over 5.15? It is kind of nice as I don't have to remotely turn off Sentry Mode before returning to the car.

 

[JeffK]

Low tech solution: What if you simply used a USB stick with more capacity...

 

[brobinson]

The single event recording when you get back to the car equals around 100mb. Seems entirely unnecessary to worry about disabling sentry first.

 

[Need]

Low tech solution: What if you simply used a USB stick with more capacity...

But I like my 16GB drives LOL.

 

[Need]

The single event recording when you get back to the car equals around 100mb. Seems entirely unnecessary to worry about disabling sentry first.

Where did you get 100mb? Yeah the actually footage would be in the 30MB 1-min file of the left repeater camera recording me coming back to the car, but the folder of the event will have 10 min of front/left/right files = 10min x 3cam x 30MB = 900MB. And then also you will have to delete them eventually even if I have a huge flash drive as they never get overwritten.

 

[GolanB]

Seems like Sentry Mode should add automatic storage management features such as first in first out recording, time-based deletion, etc. so we can get out of the business of manual data management.

In the long term, it would be nice to be able to configure it for cloud-based offloading. A couple of scenarios would justify this:

- If your car gets broken into, and the USB key is stolen, it would be nice to have access to the recorded data via the cloud.
- If your USB key encounters a failure, it would be a handy backup.

Realizing there are challenges around network connectivity, moving large file sizes around this should not be an insurmountable problem. Network connectivity is always improving, for a few minutes of recording, 4G should bear the load, and we will eventually move to 5.

Tesla may also consider placing at least one USB port available in the glove box or in a secure location, which could be locked.

 

[Need]

Yeah I like the storage management idea. I also would like to have some kind of display of the # of events recorded while I was gone. If I come back to the car and I see 1 event, okay that's just me and I can ignore it. If I see "2", I could take a look at the 2nd clip to see what triggered it. Eventually it will probably get old looking at other people getting in/out of their cars, but right now it is still fun to check. Occasionally I will see something interesting. The other day, I caught a coworker walked all around my car looking thru the window and admiring it... I caught him on all the cameras LOL.

 

[GolanB]

Yeah I like the storage management idea. I also would like to have some kind of display of the # of events recorded while I was gone. If I come back to the car and I see 1 event, okay that's just me and I can ignore it. If I see "2", I could take a look at the 2nd clip to see what triggered it. Eventually it will probably get old looking at other people getting in/out of their cars, but right now it is still fun to check. Occasionally I will see something interesting. The other day, I caught a coworker walked all around my car looking thru the window and admiring it... I caught him on all the cameras LOL.

There is no end to the number of features which could be added. With the GPU capacity of the hardware in version 3 we could theoretically leverage this for computer vision with features similar to what Google has added with their Cloud Cameras. This includes identifying known subjects, highlighting strangers and differentiating between humans, animals, and objects.