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

17.11.45 Automatic Emergency Braking

This site may earn commission on affiliate links.
Has anyone been able to test the emergency braking and make sure it functions as described? Or do you suggest we all just take their word for it?!?!

The issue of cars merging at low speeds in traffic is a big deal. I've almost been in an accidentally daily getting cut off and the car just keeps going like no one is there, at 35+ no issues but at 10 to 20 mph it's a huge issue. 134 freeway west bound at Hollywood Way on ramp merging traffic it happens every day!!!! Also my car seems to be okay with taking turns lazily and crossing over into the next lane. 134 to 101 north to Coldwater Cyn the car drives like a maniac but only north bound south bound no issues. I'd also like to see auto wipers and high beams now.
 
I'm a bit worried about this. I had to make an abrupt lane change a week ago due to a truck pulling out in front of me at highway speed. As I changed lanes, the collision warning alarm sounded, but I didn't think I was in any danger. Fortunately it was only a warning tone and not full emergency braking (since I don't yet have the AEB firmware). If the car started braking just as I was squeezing in between cars while changing lanes, I certainly would have been rear-ended.

This is a dangerous one, please please Tesla make sure this is fully tested or someone will get hurt.
 
That's good to know. I have not seen the release notes yet. Adding a steering angle sensor would help in the lane change scenario. Thanks for point this out.
They do have steering angle sensors as well as inertial and gyroscopic sensors on the ESP unit....


AEB is only meant to kick in when a collision is inevitable. But FCW is more meant to be a "watch out" warning. It will sometimes kick in when you feel like you have the situation under control.
 
Just about every mid grade and above semi-luxury vehicle has auto high beams and auto wipers. They do it without the substantial computing power and number of cameras of the Tesla. Stop making excuses for them. They work on AP1 so why can't they get them to work on AP2? They probably aren't even really trying.

I love people that say auto high beams and wipers would be easy. If so, why the heck aren't they in there yet? They only require computer vision processing to see raindrops on the windshield or detect rain on the road, and computer vision that can detect other cars at night and separate them from streetlights and other stationary lights. Yep, easy peasy. AP2 can't even read speed limit signs!

As for low priority? Tesla is screwing that up. My wife could hardly care about the AP features but is pissed she moved from a 15 year old BMW we sold for $7,000 to a $100K car and the auto wipers don't work in Seattle. For some of us, auto wipers would be one of the largest updates Tesla could make to AP2.

My gut says it's one of the hardest things they are going to do with AP2. The cameras are not focused on the windshield. Think of what kind of image detection they're going to need to do to determine it's raining by just observing the world visually.
 
Just about every mid grade and above semi-luxury vehicle has auto high beams and auto wipers. They do it without the substantial computing power and number of cameras of the Tesla. Stop making excuses for them. They work on AP1 so why can't they get them to work on AP2? They probably aren't even really trying.
Those use separate discreet sensors, which AP2 doesn't have (not sure if AP1 has them). AP1 uses Mobileye's patented techniques for camera based auto high beam, so Tesla has to figure out another way for AP2 that doesn't violate their patents.
 
Tesla vehicles don't come installed with a dedicated rain-sensing or auto-high beam sensors. They have to be programmed using the current AP cameras. As it has been mentioned above, AP1 uses the mobieye technology which is not available for AP2.

Personally i would've rather tesla put out the hardware and then upload the software later rather than just spitting out cars with absolutely no functionality at all with no possible chance for future functionality... but hey to each their own :-/

I'm sure AEB works as described, otherwise it would'nt have taken them this long for it to come out. That being said, i don't plan on testing it but i'm sure i'll see it in action eventually.
 
AP1 has the sensor. It is in MobilEye's unit.

From another post:
The rain sensor on AP1.0 cars is in the foward camera enclosure, facing the windshield. the black circle very clearly from outside. The sensor is made by Hella, and connected to 12V, ground and the Body Control Module (BCM) by Pektron, that resides in the passenger footwell, behind the carpet near the floor. The signal is LIN (Local Interconnect Network) instead of CAN.


2015-Tesla-Model-S-P85D-front-sensor.jpg

11055455_1414986392143239_625489607_n.jpg
 
Those use separate discreet sensors, which AP2 doesn't have (not sure if AP1 has them). AP1 uses Mobileye's patented techniques for camera based auto high beam, so Tesla has to figure out another way for AP2 that doesn't violate their patents.

Not sure how valid a patent is on "camera-based auto high beam" when Cadillac's have done the same thing for at least the last 50 years.
 
  • Like
Reactions: clostridium
Just about every mid grade and above semi-luxury vehicle has auto high beams and auto wipers. They do it without the substantial computing power and number of cameras of the Tesla. Stop making excuses for them. They work on AP1 so why can't they get them to work on AP2? They probably aren't even really trying.

You'll notice that I said the auto wipers would be the hard part, not auto high beams. As others have said, they used MobilEye's vision processing in AP1 and now need to make their own. The auto industry has had camera based auto high beams for at least a decade now in various platforms (not just MobilEye) so it's unquestionably possible. It's not "easy" but the world has shown it can be done and with way less processing power than Tesla has in AP2.

Auto wipers? Tesla ditched a $25 sensor based on simple physics that has been used for 20 years and on AP1 cars. They replaced it with... Nothing. They are assuming they can use the AP cameras to do this. These cameras are not focused on the windshield, so they can't see individual rain drops. So Tesla needs to develop vision processing that has never been done before, processing that will need to work like a human deciding that "it's wet out" and "how wet is it" all from a visual image. Even worse, they have less data than the human because we sit back from the windshield and can see all the glass and decide how obscured it is, while Tesla will only have data for about 1 square inch of the windisheld and can't actually focus on that. Just because it's wet out doesn't mean it's raining. We care about the windshield.

They've bitten off a huge task here with a lot of risk. I wish them well, but I have well founded doubts.

Someone with twitter should ask Elon when Tesla expects to have these features.
 
You'll notice that I said the auto wipers would be the hard part, not auto high beams. As others have said, they used MobilEye's vision processing in AP1 and now need to make their own. The auto industry has had camera based auto high beams for at least a decade now in various platforms (not just MobilEye) so it's unquestionably possible. It's not "easy" but the world has shown it can be done and with way less processing power than Tesla has in AP2.

Auto wipers? Tesla ditched a $25 sensor based on simple physics that has been used for 20 years and on AP1 cars. They replaced it with... Nothing. They are assuming they can use the AP cameras to do this. These cameras are not focused on the windshield, so they can't see individual rain drops. So Tesla needs to develop vision processing that has never been done before, processing that will need to work like a human deciding that "it's wet out" and "how wet is it" all from a visual image. Even worse, they have less data than the human because we sit back from the windshield and can see all the glass and decide how obscured it is, while Tesla will only have data for about 1 square inch of the windisheld and can't actually focus on that. Just because it's wet out doesn't mean it's raining. We care about the windshield.

They've bitten off a huge task here with a lot of risk. I wish them well, but I have well founded doubts.

Someone with twitter should ask Elon when Tesla expects to have these features.

Exactly. I love my car but I think too many people get wrapped up in the engineering challenges that Tesla takes on and miss the forest for the trees. These two features are present on many vehicles for many years. Tesla said that ours would have them. They may or may not have made some choices along the way that complicate the implementation. That all falls under the heading of "not supposed to be our problem." Don't get me wrong - the techie part of me finds it interesting what challenges they have to make it work. The problem is that this is ultimately about results not the challenge.

It's also hard not to be cynical about the recent AEB implementation. Might be coincidence but they pushed the feature out in AP2 vehicles awfully close in time to the Consumer Reports statement criticizing the lack of AEB. The broader group of auto consumers don't care about the details like all of us do on this discussion board. They want the car to work right. If Tesla wants to sell the Model 3 to the masses they need to figure this out.
 
  • Like
Reactions: EVger and Kant.Ing