Which ADAS is Really Driving? It’s Mobileye!

[JM.]

So your source is “trust me bro”.

No. Why is it that people like you don't define their claims in clear language?

https://www.nature.com/articles/s41467-024-48526-4

TL/DR; Autonomous systems are overall safer than humans but more unsafe in specific instances AT THE TIME THE ARTICLE WAS PUBLISHED. This technology is evolving daily.

Exactly, by using vague claims, some people here can be right in their own minds while being wrong overall.

Sponsored

 

[Donald Stanfield]

No. Why is it that people like you don't define their claims in clear language?



Exactly, by using vague claims, some people here can be right in their own minds while being wrong overall.

If my language is unclear perhaps you need to take a remedial class on reading comprehension. Considering the complete lack of source material backing your claims I'd start there, then perhaps an English class to learn how to cite sources. I'm not the one on here making claims, you are. Source them.

 

[godfodder0901]

So your source is “trust me bro”. Why is it that people like you never have any actual studies supporting your claims?

To be fair, it seems you used the exact same source...

 

[Donald Stanfield]

To be fair, it seems you used the exact same source...

My source is the fact that no such system is in operation without supervision. You cannot really cite a negative. Plus it's not on me to prove these systems aren't better than humans. It's on the person making the claim that they are to prove it, as human drivers are the default.

 

[JM.]

You have vaguely claimed that ADAS is not "better" than a human. Better at what? Better when? In comparison to a human in what state? Against what sort of threat, such as the ones that are not visible to the driver?

ADAS is better than a human when:
Human is tired or distracted.
Better when there are multiple problems going on, it can see all at once.
Etc.

Since ADAS literally means "driver assist" then by definition it can't be unsupervised. Your red herring doesn't work.

To be fair, it seems you used the exact same source...

Exactly, I take discussions in the same language and positions being offered by others.

 

[Donald Stanfield]

You have vaguely claimed that ADAS is not "better" than a human. Better at what? Better when? In comparison to a human in what state? Against what sort of threat, such as the ones that are not visible to the driver?

ADAS is better than a human when:
Human is tired or distracted.
Better when there are multiple problems going on, it can see all at once.
Etc.

Since ADAS literally means "driver assist" then by definition it can't be unsupervised. Your red herring doesn't work.



Exactly, I take discussions in the same language and positions being offered by others.

This doesn't mean anything. ADAS systems CANNOT operate without human supervision. Giving me little isolated cases means nothing, as you cannot operate an ADAS without human supervision. There isn't a single system good enough to operate independently, therefore they are not at human levels.

I don't see how that's such a difficult concept to grasp.

 

[JM.]

ADAS systems CANNOT operate without human supervision.

Well, it appears now you finally get what the two middle letters in ADAS mean.

Where it operates, it obviously operates well above human levels.

Where it doesn't operate at all, it's not "better" than a human.

 

[Donald Stanfield]

Well, it appears now you finally get what the two middle letters in ADAS mean.

Where it operates, it obviously operates well above human levels.

Where it doesn't operate at all, it's not "better" than a human.

Provide proof of this claim.

 

[JM.]

Hilarious. I provided examples, and empirical proof is in the thread already. Did you miss my posts or just choose to play like you did? Did you miss all the evidence or just choose to pretend that you did?

I look forward to what forms of excuse-making and strawmanning you will do when I start posting links.

Grabbing popcorn.

 

[Donald Stanfield]

Hilarious. I provided examples, and empirical proof is in the thread already. Did you miss my posts or just choose to play like you did? Did you miss all the evidence or just choose to pretend that you did?

I look forward to what forms of excuse-making and strawmanning you will do when I start posting links.

Grabbing popcorn.

Do you have studies proving these claims, studies you posted already? Let's see them. I also noticed how you didn't tag me in this response to avoid notifying me. I have seen NO proof whatsoever. Trust me bro is not proof.

This isn't even debatable; as we all know, ADAS systems cannot operate without human supervision. Since their purpose is driving, but they cannot drive without supervision, there cannot possibly be an argument that they are better.

Take something like ABS for example. ABS works better at modulating brake pressure than a human driver ever could. Knowing how it works is evidence of this, but even if you didn't know how it worked the fact that it's mandated for all new cars would be a significant clue. Not only are ADAS systems not mandated, but they are legally restricted from operating ONLY under direct human supervision because they are NOT to the level of human drivers yet.

None of this is disputable, yet here you are making unproven claims.

 

[pamalabama]

Duh. Anyone driving in a foreign country is inherently going to have greater challenges. This comment is irrelevant in the current discussion - nobody mentioned foreign driving.


A minority of people may drive like a tourist in new places (perhaps you know from first hand experience) but most people do not. I’ve traveled extensively throughout the US, driving myself and with others driving. This statement is nonsense.


Also nonsense. Ripple still occurs due to the vast majority vehicles not having ADAS, and because people change lanes in ADAS calculated space, which causes the ADAS to then slow down, creating a ripple effect. I’ve experienced it many times in my Rivian and in Teslas. In fact, human drivers can minimize the ripple by temporarily allowing the shorter distance.

except tesla does not keep a fixed distance with cars in front. You should know that. Also humans when a car merges in do slow down and even if it's more gradual still causes a ripple. Because "temporatily" allowing the shorter distance is not a permanent solution and slowing down is required.

 

[pamalabama]

Do you have studies proving these claims, studies you posted already? Let's see them. I also noticed how you didn't tag me in this response to avoid notifying me. I have seen NO proof whatsoever. Trust me bro is not proof.

This isn't even debatable; as we all know, ADAS systems cannot operate without human supervision. Since their purpose is driving, but they cannot drive without supervision, there cannot possibly be an argument that they are better.

Take something like ABS for example. ABS works better at modulating brake pressure than a human driver ever could. Knowing how it works is evidence of this, but even if you didn't know how it worked the fact that it's mandated for all new cars would be a significant clue. Not only are ADAS systems not mandated, but they are legally restricted from operating ONLY under direct human supervision because they are NOT to the level of human drivers yet.

None of this is disputable, yet here you are making unproven claims.

ABS is just one aspect of driving, just like auto cruise control systems keep more accurate distance with the car in front and reduce traffic ripple.

So are you better off using auto cruise control and steering yourself? No, because auto cruise control systems usually do not detect stationary vehicles so they have their own safety issues.

So the point is, what's wrong with using auto driving and then taking over in the few circumstances where auto driving fails?

Waymo makes weird decisions at times but in general it has surpassed human driving

 

[Zoidz]

except tesla does not keep a fixed distance with cars in front. You should know that. Also humans when a car merges in do slow down and even if it's more gradual still causes a ripple. Because "temporatily" allowing the shorter distance is not a permanent solution and slowing down is required.

Lol. In the Teslas I have driven, it does keep a fixed distance at given speed, and the distance adjusts as the speed changes. You have experienced different? Your Telsa continually changes the speed/spacing while everyone is driving at the same speed?

When cars merge in when I am driving, I do not necessarily slow down - it depends on the distance from me that they have merged in. I may choose to keep the tighter spacing depending on traffic conditions, and I have seen MANY other drivers so exactly the same.

Not all humans automatically slow down when a vehicle merges in. If the vehicle that merges in appears to be making a pass and will get back over in the right lane, I (and again, many others) will maintain speed with a shorter spacing until they return to the right lane. COMPLETELY avoiding any slowdown ripple.

When I drive I read traffic 1/4 to 1/2 mile or more ahead to observe behavior such as this by other drivers. It's called situational aware driving - ADAS does not process what a driver half a mile ahead has been doing for the past 3 minutes. Human Driver +1, ADAS 0.

 

[Donald Stanfield]

ABS is just one aspect of driving, just like auto cruise control systems keep more accurate distance with the car in front and reduce traffic ripple.

So are you better off using auto cruise control and steering yourself? No, because auto cruise control systems usually do not detect stationary vehicles so they have their own safety issues.

So the point is, what's wrong with using auto driving and then taking over in the few circumstances where auto driving fails?

Waymo makes weird decisions at times but in general it has surpassed human driving

Nothing, until you claim it’s equivalent to a human driving.

 

[captainjp]

Waymo makes weird decisions at times but in general it has surpassed human driving

Like getting stuck in roundabouts?

Sponsored