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

A cautionary autopilot video

Do you use autopilot on secondary roads (not highways)?


  • Total voters
    208
  • Poll closed .
This site may earn commission on affiliate links.
Again, what manufacturers have AEB crash prevention systems? None. They all say they only reduce impact.

At highway speeds that's correct.
At lower speeds that isn't correct.

The Subaru Eyesight system says that it will stop (without hitting the object) at up to a speed differential of just over 31mph. Which is kinda funny considering the IIHS used a speed of 31mph toward a stationary vehicle.

The Tesla (both AP1 with the Model S and AP2.5 with the Model 3) both hit the stationary target.

This is a good article on AEB systems. It's really unfortunate that the NHTSA sets such a low bar for them.

Automated-Emergency-Braking Systems Don't Always Work - Test Results
 
Last edited:
At highway speeds that's correct.
At lower speeds that isn't correct.

The Subaru Eyesight system says that it will stop (without hitting the object) at up to a speed differential of just over 31mph. Which is kinda funny considering the IIHS used a speed of 31mph toward a stationary vehicle.

Please SHOW US WHERE Subaru makes that claim.

In Subaru's 2019 Forester EyeSight Owners Manual it does not say "...it will stop (without hitting the object) at up to a speed differential of just over 31mph."

Page 12: "Pre-Collision Braking System This function uses a following distance warning feature to warn the driver to take evasive action when there is the possibility of a collision with a vehicle or obstacle in front of you. If the driver does not take evasive action, the brakes are applied automatically to help reduce vehicle collision damage or, if possible, help prevent a collision."

Page 26 "When the difference in speed with the obstacle in front is the following figure*1 or more, it may not be possible to avoid a collision. Even if the speed difference is the following figure*1 or less... "
*1: For vehicles: approximately 30 MPH (50 km/h)

Subaru Cars, Sedans, SUVs | Subaru of America


The Tesla (both AP1 with the Model S and AP2.5 with the Model 3) both hit the stationary target.

They used a 2016 Model S with v7.1 and a 2018 Model 3 v8.1. They should have atleast used v8.1 for the Model S which was the latest version available at time of testing. For the Model 3, v8.2, came out shortly after the article and fixed the Consumer Reports braking issue.
 
  • Informative
Reactions: StealthP3D
I doubt they would get anything. They crossed the street outside of the crosswalk. In some jurisdictions, it's against the law.

Burglary is illegal as well. But if you shoot a fleeing burglar, they or their estate can sue and will likely win.

Jaywalking is illegal in almost all jurisdictions but the pedestrian always has the right-of-way. I thought everyone knew this.
 
Please SHOW US WHERE Subaru makes that claim.

In Subaru's 2019 Forester EyeSight Owners Manual it does not say "...it will stop (without hitting the object) at up to a speed differential of just over 31mph."

Page 12: "Pre-Collision Braking System This function uses a following distance warning feature to warn the driver to take evasive action when there is the possibility of a collision with a vehicle or obstacle in front of you. If the driver does not take evasive action, the brakes are applied automatically to help reduce vehicle collision damage or, if possible, help prevent a collision."

Page 26 "When the difference in speed with the obstacle in front is the following figure*1 or more, it may not be possible to avoid a collision. Even if the speed difference is the following figure*1 or less... "
*1: For vehicles: approximately 30 MPH (50 km/h)

Subaru Cars, Sedans, SUVs | Subaru of America

On Page 26 that you quoted.

When the difference in speed with the obstacle in front is the 50km/h (roughly 31mph) or more, it may not be possible to avoid a collision.

The 50km/h figure is for cars, and 35km/h is for pedestrians.

That to me means in most cases (not all as driving is full of unpredictable things) that the Subaru will stop without hitting a vehicle if the speed differential is 50km/h or less. Or up to 35km/h if it's a pedestrian. I imagine it's lower with pedestrians as they are smaller objects.

I think its worth pointing out that in the Car and Driver article that the Subaru was actually successful braking without hitting the dummy vehicle well past that 31mph mark. So I think Subaru is being overly cautious with their numbers.

That's also not the only reference information on it. Subaru is one of the only companies I know of that goes in depth as to how their system operates, and what to expect from it.
 
They used a 2016 Model S with v7.1 and a 2018 Model 3 v8.1. They should have atleast used v8.1 for the Model S which was the latest version available at time of testing. For the Model 3, v8.2, came out shortly after the article and fixed the Consumer Reports braking issue.

I agree, and that actually really bugged me at the time. It did because it made it impossible to compare the difference between AP1 in the Model S, and AP2.5 in the Model 3.

I wanted to see how much of the false braking (in TACC/AP) was due to the changes in Firmware 8.1 or the difference Hardware. Lots of people with AP1 are convinced that Firmware 8.0+ greatly increased false braking events under TACC/AP. I didn't experience that with my AP Model S, but I did have more false FCW alerts with Firmware 8.
 
The 50km/h figure is for cars, and 35km/h is for pedestrians.

That to me means in most cases (not all as driving is full of unpredictable things) that the Subaru will stop without hitting a vehicle if the speed differential is 50km/h or less. Or up to 35km/h if it's a pedestrian. I imagine it's lower with pedestrians as they are smaller objects.

35 kph is only 21 mph. The OP's car was going faster than that in the video.
 
The discussion of the Subaru AEB system was WAY OUTSIDE the discussion about the OP's car.

Yeah, but the Subaru might have run over the pedestrian in the same situation. I was following a Subaru tonight that was struggling to make it up some of the slippery hills on the highway. It was 33 degrees F and snowing up a storm. The P3D was tracking true and straight, only slipping when I floored it to have some fun since I was so I was so bored going 23 mph in a 50. Subarus have to be some of the lowest tech vehicles on the road. And the most over-rated in the snow. It is the official car of the US Ski Team which is just good marketing.
 
Last edited:
  • Funny
Reactions: AlanSubie4Life
It would be nice if @rmacey would answer our questions about the FCW sounding, and what sensitivity it was set to, but he probably has better things to do.

My guess is that it did not alert with FCW though.

The OP's car was going faster than that in the video.

I would agree with that. I looked at the video and it looked like it was braking for about 1.5 seconds. It was wet, but not too wet, so using the 125ft 60-0 braking distance with MXM4s (don't know what he was driving), that works out to 0.96g, so that would suggest initial velocity was about 0.96*9.81m/s^2*1.5sec = 32mph. That could be off if the braking was not optimal or if the time was a little shorter (it's an estimate on my part). But I'd guess 28-35mph.

In all those cases we're relying on the HW2.5 computer running a neural network that's supposed to detect pedestrians.

Really nit-picking here. Sorry. :)
It's not really *supposed* to detect pedestrians. I mean, it will sometimes detect them, it definitely tries! But Tesla specifically says it won't detect them sometimes. So it's perhaps a little strong to say it is supposed to detect them, since some people would take that to mean they should "expect" the car to detect them, which certainly we should not!

Definitely it's nice to have the FCW (I set mine to maximum sensitivity, as I want to maximize the time I have to brake, to avoid someone piling into me), and it goes off fairly often (I greatly appreciate that it goes off when someone a couple hundred feet ahead hits the brakes hard). But it's just supplemental insurance. I only use TACC & AP on the freeway and I'm almost always fighting with it; Autosteer really has no idea how to drive properly. But I haven't taken a long freeway road trip yet (I think it would go a lot more smoothly then, and would be very helpful).
 
I think its worth pointing out that in the Car and Driver article that the Subaru was actually successful braking without hitting the dummy vehicle well past that 31mph mark. So I think Subaru is being overly cautious with their numbers.

No it was not successful. If you read the article, Subaru hit the dummy vehicle the hardest.

"In our stationary-vehicle test, the Impreza's first run at 50 mph resulted in the hardest hit of the day, punting the inflatable target at 30 mph."
 
  • Informative
Reactions: StealthP3D
It's not really *supposed* to detect pedestrians. I mean, it will sometimes detect them, it definitely tries! But Tesla specifically says it won't detect them sometimes. So it's perhaps a little strong to say it is supposed to detect them, since some people would take that to mean they should "expect" the car to detect them, which certainly we should not!

True, obviously it will not detect them every time. That said, tonight I was driving on a dark, rural road. I had my auto-high beams on and since there was no traffic (and no reflectors or anything else to fool it) and so my high beams were illuminated. As soon as a pedestrian dressed all in dark, non-reflective clothing and walking on the shoulder came into my view (illuminated by the highbeams) they automatically dimmed. The only thing I can figure is the pedestrian was detected on the shoulder and the beams were lowered so as to not blind them.

Very nice touch Tesla! And I'll note that the Auto High Beams have been working very well indeed on the dark rural roads I drive so often. My only complaint is they are still a little slow to go back to high after the oncomming traffiic has passed. It takes about 1.5-2 seconds.
 
No it was not successful. If you read the article, Subaru hit the dummy vehicle the hardest.

"In our stationary-vehicle test, the Impreza's first run at 50 mph resulted in the hardest hit of the day, punting the inflatable target at 30 mph."

That was the first one done at 50mph, and then the second attempt brought it down to 12mph.

If you look at the chart for you'll see that they never even tested the other cars at 50mph. They only tested the Subaru because it successfully avoided the accident at 45mph.

"The Subaru Impreza—the least expensive car of the four, with a stereo-camera system that eschews the usual radar sensor—still prevented a collision at 45 mph, a higher speed than any other car here"

Keep in mind there are a lot of weaknesses to a Stereo only based solution outside of that specific test. So I'm not saying the Subaru system is the best system overall. I'm simply saying it's pretty impressive for the price.
 
Keep in mind there are a lot of weaknesses to a Stereo only based solution outside of that specific test. So I'm not saying the Subaru system is the best system overall. I'm simply saying it's pretty impressive for the price.

Are you a price-conscious shopper when buying cars that could mean the difference between life and death of your loved ones? If we can save a few bucks, maybe we'll be able to afford a big screen for the bedroom?
 
Are you a price-conscious shopper when buying cars that could mean the difference between life and death of your loved ones? If we can save a few bucks, maybe we'll be able to afford a big screen for the bedroom?

There is no perfect system on the market.

There are strengths and weaknesses of each approach. The Tesla approach coupled with HW3 is going to be pretty strong as we have both Radar, and Multiple front facing Cameras. The HW3 will add quite a bit more processing power than we have now, and should considerably increase both the detection accuracy and the number of objects it can detect.

The advantage of the Subaru system is it's lower cost, and gets more people into cars with a much higher level of crash avoidance/mitigation than used to be available at that price point. Oh, and it also has a proper blindspot monitoring system with the indication in the mirrors. It also has rear cross traffic detection for when you backup.

For most driving situations the Subaru system is great. I made sure my mom got the EyeSight system when she bought a Subaru Outback a few years ago. She's considerably more budget conscious than I am, but I still wanted her to get things like AEB, FCW, Blindspot Monitoring, ACC, etc.

Much of the advantages a radar would have added don't benefit her. She isn't going to be use adaptive cruise control (she doesn't drive on highways), and she doesn't drive in visually limited situations where radar is better at tracking the vehicle in front. So the vehicle fits her needs just fine, and eventually my brother will take the car as she doesn't drive it that much. He has a family so he should be driving a car with those features, and it's also the convenience features as well.

He's too much of a cheap skate to get that kind of stuff himself. He's too busy saving money for things like college tuition for his kids, and other ridiculous things like that. :p

Where I don't have kids so I have a lot more freedom, and I generally have much higher expectations with what I buy. I buy more for convenience, speed, excitement than safety. If I wanted something really safe I'd probably get a Volvo XC90. It's regarded as the safest car in the world by the IIHS, and others. But, it's far too boring/slow for me and isn't an EV.