Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register
This site may earn commission on affiliate links.
Currently I have faster reaction time than FSD Beta (assuming I am paying attention and see the relevant object/incident - where FSD theoretically has an advantage), neglecting any anticipatory ability (which would give me additional advantage).
Here is an experiment you can try.

Based on a 2020 article, A group from the UK analyzed 10 viral videos of Tesla’s Automatic Emergency Braking (AEB) in action, breaking it down frame-by-frame to estimate the average response time of AEB to be 0.3 seconds (300 milliseconds).

This would seem to support your observation, as average human reaction time is about 273ms.
 
  • Informative
Reactions: Phlier
Here is an experiment you can try.

Based on a 2020 article, A group from the UK analyzed 10 viral videos of Tesla’s Automatic Emergency Braking (AEB) in action, breaking it down frame-by-frame to estimate the average response time of AEB to be 0.3 seconds (300 milliseconds).

This would seem to support your observation, as average human reaction time is about 273ms.
I guess I need to play more video games, my reaction time is 230ms. That other game is garbage though and I don’t know how they assessed AEB reaction time; there’s no evidence for their assertion I could see linked.

But for the type of scenario I am discussing, it’s not really a question of reaction time in this traditional sense. This wouldn’t be an AEB event. It’s a scenario where a reaction is required early to make acceleration and acceleration changes lower and more tolerable and still fit the maneuver within the required distance. It’s a perception and reaction latency problem. AEB events usually represent driver failure (though there are exceptions of course).
 
Here is an experiment you can try.

Based on a 2020 article, A group from the UK analyzed 10 viral videos of Tesla’s Automatic Emergency Braking (AEB) in action, breaking it down frame-by-frame to estimate the average response time of AEB to be 0.3 seconds (300 milliseconds).

This would seem to support your observation, as average human reaction time is about 273ms.

I only attempted once. 186ms

I can also catch the dropping dollar bill, much to the frustration of my kids :D

As far as how far we can see, you really need to define the parameters. After all, I can see the Andromeda galaxy with my bare eyes.
 
I guess I need to play more video games, my reaction time is 230ms. That other game is garbage though and I don’t know how they assessed AEB reaction time; there’s no evidence for their assertion I could see linked.

But for the type of scenario I am discussing, it’s not really a question of reaction time in this traditional sense. This wouldn’t be an AEB event. It’s a scenario where a reaction is required early to make acceleration and acceleration changes lower and more tolerable and still fit the maneuver within the required distance. It’s a perception and reaction latency problem. AEB events usually represent driver failure (though there are exceptions of course).
Tesla has about 820 feet to play with. 75MPH is about 110 feet per second, so at 75MPH that's just over 7 seconds of reaction time (assuming the system is reacting at max distance of the camera).

According to this site, it would take 356 feet to stop at that speed, but that's hard braking. Even if it stopped in the full 820 feet, that's still pretty solid braking power. Obviously if the cameras could see farther, like Waymo can see 1600 feet, it could brake more gently assuming it reacted the moment it noticed a stopped car at max camera distance.
 
  • Informative
Reactions: Phlier
I only attempted once. 186ms

I can also catch the dropping dollar bill, much to the frustration of my kids :D

As far as how far we can see, you really need to define the parameters. After all, I can see the Andromeda galaxy with my bare eyes.
I did qualify it in another post, the average human eye can see a candle flame at about 8400 feet. Obviously the larger the object, the farther away we can observe it. Since we're talking about an object roughly the size of a car, I'd think it reasonable we can observe it farther out than 8400 feet, but wouldn't be able to make out details clearly.

Actually this article gives good data on the human eye. "Human-scale objects are resolvable as extended objects from a distance of just under 2 miles (3 km). For example, at that distance, we would just be able to make out two distinct headlights on a car."
 
well we have some news on 10.12.1 at least . . .
the one guy on TeslaFi holding out installing the latest beta has finally got around to installing it at last.
Maybe that means they'll carry on the rollout now, but he may need to post a youtube video of how smooth it is.
🤣
About friggin' time!!! There's nothing more infuriating than being a beta tester frothing at the mouth for 10.12.1 and seeing that the lone M3 tester on Teslafi got it days ago and refused to install it...
 
You never know, their car could have been in the shop. (Like Chuck's Model 3 that was in an accident.)
Or he could have been in hospital getting cardiac stents, or his mother-in-law was visiting from Japan, or.........maybe he lives in Tibet and lost internet service in an ice storm........you never know......
 
Anticipating is a pretty tall order for FSD or any AI system. Not only do humans often have trouble but it's depending on regional driving styles and even time of day. I wouldn't expect that from any driving system any time soon.
And as far as FSD is concerned to me it doesn't matter. Does FSD get me from point A to point B without intervention and without causing an incident? That is my holy grail. I don't care if it gets cutoff, honked at, pisses off aggressive drivers, etc. Just as long as the granny's are happy, I'm happy.
 
And as far as FSD is concerned to me it doesn't matter. Does FSD get me from point A to point B without intervention and without causing an incident? That is my holy grail. I don't care if it gets cutoff, honked at, pisses off aggressive drivers, etc. Just as long as the granny's are happy, I'm happy.
It’s possible that no anticipation ability is mutually exclusive of safety level 10x that of a human driver. I’m really not sure; it’s actually hard to say because we don’t know how many accidents are avoided due to this capability. Clearly FSD could calculate trajectories and anticipate in that sense, so that is good, but further “next level” anticipation might be harder and I don’t know what additional level of safety it provides.
 
What in the world happened to the rollout of 10.12? Only 3 on TeslaFi and that was almost a week ago.
I know all the shillers got it and they’re all gloating but come on.

Ski

Nothing unusual based on the last two upgrades. Plus it is better they address the initial problems to a small number of testers.
The shillers are also identifying problems and reporting them to Tesla.

Release History.png


Wiki - FSD Beta release history
 
  • Like
Reactions: Phlier
This again? Stop listening to him. It's not that hard. Only believe what the company says. No where in your Tesla ordering experience did it say you'd have it by a specific date. Elon says something optimistic, and I am hopeful, but I NEVER expect what he says is actually going to happen.

Do you believe politicians too?
Sometimes the employees echo what Elon says without mentioning the name Elon
 
  • Like
Reactions: 2101Guy