The competition has completely crushed Tesla FSD (AP).

[alexgr]

One more funny article out of the series "publish fabricated story about Tesla and keep doing it until everyone is convinced it's true"

How Tesla, BMW, Ford, GM and Mercedes driver assist systems compare | TechCrunch

We tested vehicles from BMW, Tesla, GM, Ford and Mercedes equipped with advanced driver assistance systems. Here's how they performed.

techcrunch.com

 

[SidetrackedSue]

This is not true.

Overriding Tesla’s Autosteer functionality requires a firm grip and a sharp turn of the wheel. This is unnerving for the driver and uncomfortable for the passengers.

It is also not true that the article was testing FSD(AP) as you state in your post title. As tesla has chosen confusing and misleading names for their products it is incumbent on the rest of us to be accurate and not give tesla any slack for their deliberate deceptions.

The articla was testing FSDb v11.4.2 which is quite old now but that's moot, FSDb has not improved dramatically since then, in fact in many ways is worse now.

The article was not testing AP as you stated in your title.

 

[alexgr]

This is not true.



It is also not true that the article was testing FSD(AP) as you state in your post title. As tesla has chosen confusing and misleading names for their products it is incumbent on the rest of us to be accurate and not give tesla any slack for their deliberate deceptions.

The articla was testing FSDb v11.4.2 which is quite old now but that's moot, FSDb has not improved dramatically since then, in fact in many ways is worse now.

The article was not testing AP as you stated in your title.

I put FSD (AP) in the title as Tesla's FSD is nearly identical to AP on highway with exception of lane changes. I doubt TC people really tested anything, frankly.

 

[JulienW]

.... as Tesla's FSD is nearly identical to AP on highway with exception of lane changes.....

No it is NOT. It is completely different code.

 

[alexgr]

No it is NOT. It is completely different code.

I drive both. I don't know about the code, but from the consumer perspective I see no differences.

 

[Supcom]

I drive both. I don't know about the code, but from the consumer perspective I see no differences.

I drive both and it's night and day difference.

 

[SidetrackedSue]

And speaking of night and day, AP is free and FSD is $$$$$.

 

[alexgr]

I drive both and it's night and day difference.

What is different for you in highway driving between the two? We should also remember that until recently, FSD used AP code on highway.

 

[Dewg]

This is not true.



It is also not true that the article was testing FSD(AP) as you state in your post title. As tesla has chosen confusing and misleading names for their products it is incumbent on the rest of us to be accurate and not give tesla any slack for their deliberate deceptions.

The articla was testing FSDb v11.4.2 which is quite old now but that's moot, FSDb has not improved dramatically since then, in fact in many ways is worse now.

The article was not testing AP as you stated in your title.

@DirtyT3sla did a video on disabling ADAS. It was a finger on the wheel. The car wanted to turn and he just put his finger on the wheel and it disabled. I hate when media spreads disinformation.

 

[alexgr]

@DirtyT3sla did a video on disabling ADAS. It was a finger on the wheel. The car wanted to turn and he just put his finger on the wheel and it disabled. I hate when media spreads disinformation.

My point of the post exactly! There are at least two more normal ways to disable the AP/FSD but that does not stop the author from saying this: "Overriding Tesla’s Autosteer functionality requires a firm grip and a sharp turn of the wheel." Also, when he writes: "I found the FSD beta software frequently confused on urban and rural streets, often slowing down excessively ahead of curves and behaving erratically through intersections, especially on roundabouts." I wonder which other system did he even dare to test on roundabouts. So much for calling that a test...

 

[JB47394]

@DirtyT3sla did a video on disabling ADAS. It was a finger on the wheel. The car wanted to turn and he just put his finger on the wheel and it disabled. I hate when media spreads disinformation.

Cars very definitely vary in the amount of pressure it takes to disable ADAS. Mine requires so much pressure that when I break the lock, I'll overcorrect and swerve a bit. I've seen drivers demonstrate the one-finger-disengage technique and I absolutely cannot do that in my car. Because of the behavior I get, I rely on slapping up the right stalk to disable ADAS. In an emergency, I can still crank the wheel around because I'm undoubtedly trying to swerve anyway.

The variability of pressure across cars may explain the many different experiences that people have with defeating nags, receiving strikes and so on.

 

[Dewg]

Cars very definitely vary in the amount of pressure it takes to disable ADAS. Mine requires so much pressure that when I break the lock, I'll overcorrect and swerve a bit. I've seen drivers demonstrate the one-finger-disengage technique and I absolutely cannot do that in my car. Because of the behavior I get, I rely on slapping up the right stalk to disable ADAS. In an emergency, I can still crank the wheel around because I'm undoubtedly trying to swerve anyway.

The variability of pressure across cars may explain the many different experiences that people have with defeating nags, receiving strikes and so on.

That's a bummer. Have you had service take a look at your torque sensor? Perhaps they can adjust it or perhaps it needs to be replaced?

 

[JB47394]

That's a bummer. Have you had service take a look at your torque sensor? Perhaps they can adjust it or perhaps it needs to be replaced?

I never had service look at it. But I did just go for a test drive to make sure it was still the way I remember.

Magically, I can now do the one-finger-disable, and taking over steering requires only a light touch. It's very natural now, whereas when I got the car two years ago it very definitely required a lot of force. So much force that I hadn't tried taking over steering directly since then. So either there's a breaking-in period on the sensor or the pressure can be changed through software updates. Or it's just sloppy and the next time I drive it will require a lot of force. shrug

 

[Dewg]

I never had service look at it. But I did just go for a test drive to make sure it was still the way I remember.

Magically, I can now do the one-finger-disable, and taking over steering requires only a light touch. It's very natural now, whereas when I got the car two years ago it very definitely required a lot of force. So much force that I hadn't tried taking over steering directly since then. So either there's a breaking-in period on the sensor or the pressure can be changed through software updates. Or it's just sloppy and the next time I drive it will require a lot of force. shrug

As a reference, I just got back from a Vegas trip - 250 miles 1-way. One (1) nag the entire time. Welcome to how it's supposed to work.

 

[DBeckwith]

My observation is that it requires more torque to disengage when it thinks it knows better than me. It doesn't bother me but some of my passengers do complain. Yet, they are the same passengers you would characterize as back seat drivers without Auto Pilot.

 

[Supcom]

As a reference, I just got back from a Vegas trip - 250 miles 1-way. One (1) nag the entire time. Welcome to how it's supposed to work.

You must have been asleep most of the way and missed them.

 

[Dewg]

You must have been asleep most of the way and missed them.

To clarify, I keep my hands on the wheel and apply slight torque the entire time. There was one long straight portion in the desert that I saw the nag.

 

[EVNow]

@DirtyT3sla did a video on disabling ADAS. It was a finger on the wheel. The car wanted to turn and he just put his finger on the wheel and it disabled. I hate when media spreads disinformation.

One reason AP/FSD uniformly gets a bad rap in the media is ... there is some learning curve to AP (esp. FSD). Media people tend to "test" in a few hours - invariably the "tests" fail.

Its like a newbie who is touching a violin for the first time saying the violin is useless and doesn't actually sound good.

 

[Fishstyx]

I just used up the 3-month FSD trial on a new Tesla (2nd Tesla, I've been an AP user for a year) and I find the article pretty close to the truth. Some of it is sensationalized, but in general, I found the same FSD behaviors in my experience. As for "learning curve", I LOL'd at that. I think what you mean is simply a tolerance for FSD's behaviors with range from quirky/annoying to dangerous to yourself and others. There is no 'learning' for the driver since FSD mode is either on or off, and when it's on the only option the drive has is to override controls which shuts it off.

Also, I call bunk on the statement the "just put a finger on the wheel" to override statement. That assertion is obviously untrue because the only way the system knows you have your hands on the wheel is by sensing torque below the level of override. If it were too sensitive, you'd be disengaging AP/FSD frequently just trying to satisfy the nag. I have a '23 MYLR and '23 M3P and the M3P requires about twice the torque to satisfy the nag compared to my MY and (obviously) more than that to disable AP/FSD. Due to the risk of causing upset at high speed when torquing the wheel to override, I always use brake or stalk and only wheel override to keep FSD from doing something dumb or dangerous.

 

[JB47394]

As for "learning curve", I LOL'd at that

The learning curve is understanding what the system is going to do in a given circumstance so that you can know whether you want to let it do its thing. After learning what the system can do, I tend to just rely on it for long stretches of roads where its lane-keeping makes the drive less demanding. I tend not to use it in areas where it has to negotiate with other drivers because it does not play well with others.

Also, I call bunk on the statement the "just put a finger on the wheel" to override statement.

As of yesterday I would have agreed with you, and I also long ago learned to slap the stalk up to disengage. Today, I find that my car is in the sweet spot where I can maintain torque on the steering wheel to placate the nag while still being able to disengage the system with the pressure of a single finger (or just a little more torque). It's firm pressure, but not at all uncomfortable.

I think that cars vary in the amount of pressure needed to override the torque sensor which is why different people can emphatically support different positions; they have the experience in their own car to back them up. Your own experience demonstrates how different cars can be.

As Dewg suggests, you may want to take your Model 3 to a service center and see if they can adjust the torque sensor.