Have this in my BYD. It's a pain in the hole TBH (or at least the BYD implementation, acknowledging that Tesla will probably do a better job). Sets off the alarm if there's basically anything at all left on a seat (I left a <1kg package last week - it went off), and also alarms if you're just sitting in the car, even if the car is switched on. Worst of all, it resets every time so if you turn it off (say, you don't have kids!) it's back on next time - even though it's not a mandatory safety system.
Have heard about this on the BYD as my brother mentioned it. Basically can’t leave someone sitting in the car and walk into the shop if you have the key, Tesla is very similar in that regard too in fairness.
Implementation of how the software detects a person may be key.
When I am getting out of the cupras along with the 'goodbye, have you got your belongings' message, I get an option to turn off interior monitoring at that stage.
If it's only me and I'm leaving the dogs in the car going into a shop or something i select this.
So it's easily overcome if they put the proper thought into it.
That’s the same with Tesla. You can leave your phone in the car or you can turn on camp/dog mode.
The thing about people leaving kids in cars though, is that this won’t fix that. The pop up alert on the Cupra may have some chance of making a person re-check but in these cases, the people driving have some other issue that makes them forget the child.
I don’t know how it can be implemented perfectly so hopefully someone smarter than me is sorting it out 😂
No i was just saying it solves the problem of alarm going off if you dont want it to, response to the BYD comment about items left on seat being mistaken as occupancy.
As for leaving kids in cars, that's a bit bigger than we'll solve here!
Air conditioning remotely and while the car is switched off is a great feature of EVs overall though, both hot and cold.
”Morgan Stanley number cruncher Adam Jonas reckons Tesla’s 12.5 per cent first-quarter gross auto margin, its lowest level in a dozen years, is lower than almost any big carmaker. It would point to a loss-making operating margin when the dealer cut that Tesla enjoys is stripped out. Unlike traditional carmakers, Tesla sells directly to customers, cutting out the middleman.”
Stock up again.
All adds to safety, it's not the same but my bZ4X reminds me to check the back seats when I turn it off.
The problem with that kind of persistent/repeated message is that your brain will quickly learn to ignore it. A dynamic properly implemented dynamic warning is much more likely to be successful.
Testing of autonoumous fsd is going on at present in Austin with something like 300 MY (with driver for safety). They are clocking up miles driven in order to eek out edge cases, "long tail" of unusual situations. They seem to be very strongly aiming to go live with 10 to 20 fully autonomous MY cars for June 1st but prehaps really aiming for before June 30th.
There will be a remote control room, but if I understand it right, this will be to take over where the car decides it needs help rather than as an intervention override.
Cybercab will be in production late this year or quite early next year.
The expectation is that cybercab will start contributing profit during 2026, as it rolls out to more cities in volume.
Fsd seems to be getting rolled out in China currently, possibly starting in India soon, possibly accepted in Netherlands in May/June and rest of EU soon after.
We wait and see, som of us more optimistic than the average contributer on here.
FSD coming to a road near you Real Soon Now™ 🤣
Real Soon! Within the next 100 years! Funding Secured!
promising
Tick, tick, tick.
We won't need to wait 100 years to see. More like a couple of months.
Couple of months… lol
isn’t it happening in June? That’s 5 weeks away.
Tick tick tick indeed.
June, yes. But I won't tell you what year or decade!
I think your hat is safe. 🤣
if this is meant to be camera based it’s never going to work. Having used Volvo and Tesla it’s not really going to work unless you use radar which I think Tesla doesn’t have..
You need radar and lidar and camera.
If I had to pick 2 of 3 I'd pick the first two and leave out camera as it's the least reliable input. What does camera have going for it? Oh that's right… it's the cheapest by far!
musk is refining photo film to be the best in the world, while ignoring the move to smartphones
Far too simplistic, there are a lot of visual inputs that are relied upon, signage, lights. regardless of your views I doubt if there are any automobile manufacturers that are relying on lidar and radar alone without cameras to finalise FSD on a mass market level
Humans don't have lidar, yet they manage. At distance over 10 m, we guage distance by visual only, not 3D.
I think our eyes can process a lot more data than a 5MP camera.
Can humans gauge distances? -yes, they can. Are they more accurate than a simple laser range finder? -no, they are not.
Because something is possible using humans, doesn’t mean it’s best approach for us to currently try and implement.
The technology currently isn’t there. If we are closer to solving the problem using laser and radar we should use that instead for now. Waymo have a better self driving experience using these technologies than what Tesla are offering.
There is a reason we put wheels on cars first instead of trying to put them on legs.
this is such a lazy argument - on the move, people are walking at 5-10kmph at most, and almost all footpath obstacles are obvious, usually fixed, and rarely if ever come at us fast enough that we can’t avoid, or maybe even killed.
Doing that at vehicle speeds, on roads where everyone is going the same speeds, is so utterly different.
What's lazy? Humans using only eyes while driving?
We build a sufficient model of environment to plot a safe course. We depend on orderly behaviour enforced by rules. We can probably handle up to two things diverging from normal expected behaviour at one time. We employ some preemptive what-if caution, slowing down when something is suspect.
I don't see why a visual system would be fundamentally incapable of reproducing this kind of capability with suitable processing.
You're missing the point. The use of eyes is more than just 3D. Its our eyelids(cleaning), its eye movement, its head movement, its hand-eye coordination, its the brain behind the eyes.
3D is a very small part of it.
You are over simplifying the task at hand. One or two forward facing static cameras cannot adequately compensate for all the characteristics of human eyes and movement to cover the edge cases and its the edge cases are the problem.
the difference is the speed ffs! There’s almost zero safety risk when walking in public, and the main risk is being hit by vehicles.
All of your arguments here for months now ignore the safety-first element of vehicle regulations and concentrate only on the assumption that Elon has the answers to all camera limitations.
No regulator will sign off on a camera-only system for a driverless car ( and the US as we know from the Cybertruck, has no pedestrian safety regulations to begin with 🤪).
TLDR: I’m don’t doubt camera-only can get close to a workable technological solution, but they’ll never be approved while a better road safety solution exists.
The UNECE regulations are quite clear on meeting operational performance goals and are agnostic on the nature of the systems used to meet those requirements. It would be most unusual at this point in the process for the working groups handling the topic to throw away all their previous work and go down the route of a prescriptive technology stack.
