Yes, but in that respect it's the opposite of what was mentioned earlier; autonomous vehicles being trained to be more aggressive and more follow the social rules (standardized human behaviour) than the written rules.
Self driving cars… no thanks.
Re: Self driving cars… no thanks.
Re: Self driving cars… no thanks.
It could, if done less than well; if perfected it may offer something worthwhile.rareposter wrote: ↑7 Jan 2025, 4:02pm The problem with this approach... is that it leads to the driver concentrating less. Or when the system does give a warning, it's dismissed as another false positive, like the other 12 times the car "saw" a bin, a tree, a fox, a shadow and misidentified it.*
As a general rule, the more automation you put into a task, the less the human operator will concentrate on it. That works both ways - it's a benefit because sometimes people are fallible and a computer can monitor something 24/7 without ever needing a coffee break but it becomes a downside when people expect the computer to sort everything out and it can't. It's called Out Of The Loop - removing a driver from either the physical control loop (the feel of the steering wheel and pedals) and/or the cognitive control loop which is loss of situational awareness either because they are looking away from the driving scene during automation (assuming the computer will sort everything out) or because of boredom and mind-wandering.
...
Does automating headlight and wiper operation improve safety? I've previously questioned that, I wonder if night vision displays as some Lexus, Mercedes and others use should be as ubiquitous as anti-lock braking.
-
- Posts: 3590
- Joined: 11 Jan 2007, 12:20pm
Re: Self driving cars… no thanks.
The Yangwang U9 a self-jumping car from BYD.
I've had two accidents when I've stopped and the driver behind couldn't, it's not a good idea to test the reaction of drivers if you don't have to.rareposter wrote: ↑7 Jan 2025, 4:02pm *There's a well-known video of a Tesla (them again!) in full self-drive mode going into an underpass and the car gets very confused by the shadows, lights, radar return off the tunnel etc and simply does what it thinks is the right answer - slams the brakes on. Kind of a mess.
https://youtu.be/E531GxfEoB8?si=uiViHbWPrF1coVaR
People make assumptions about the braking performance of other cars. A guy at work owned an Austin 7 with cable operated brakes, he said it was a PITA to drive in the rush hour because every time he dropped back to allow himself enough stopping distance, the car behind just overtook and jumped in to fill the gap.
“I'm not upset that you lied to me, I'm upset that from now on I can't believe you.”
― Friedrich Nietzsche
― Friedrich Nietzsche
-
- Posts: 8269
- Joined: 7 Mar 2009, 3:31pm
Re: Self driving cars… no thanks.
Quite. I have often seen passengers get restive when a more cautious driver fails to push forward hard enough.mjr wrote: ↑5 Jan 2025, 10:16pmNot explicitly object but they vote with their custom and hire the faster human-driven taxis.Mike Sales wrote: ↑5 Jan 2025, 9:52pmBmblb[i wrote:[/i]zzz post_id=1887371 time=1736112560 user_id=25113]
Unacceptably to some of the motorists behind them, that is. I doubt their passengers would actually mind.That sounds as if it is the passengers who object.: When robotaxis obey law, they don't go fast enough to compete successfully with Uber, so Google execs ordered engineers to ignore laws."
If the robots have to disobey traffic law in order to conform to normal monkey driving, the conclusion is clear.
It's the same the whole world over
It's the poor what gets the blame
It's the rich what gets the pleasure
Isn't it a blooming shame?
It's the poor what gets the blame
It's the rich what gets the pleasure
Isn't it a blooming shame?
Re: Self driving cars… no thanks.
"The Yangwang U9 a self-jumping car from BYD".
With the state of the roads in Oxfordshire, it would never touch the ground!
With the state of the roads in Oxfordshire, it would never touch the ground!
-
- Posts: 2542
- Joined: 7 Jul 2008, 12:52am
Re: Self driving cars… no thanks.
This is exactly what any driver should do when they cannot see the way ahead. A common failing for human drivers is failing to stop when they can't see where they are going - so we end up with multiple pile ups on foggy motorways.rareposter wrote: ↑7 Jan 2025, 4:02pm
*There's a well-known video of a Tesla (them again!) in full self-drive mode going into an underpass and the car gets very confused by the shadows, lights, radar return off the tunnel etc and simply does what it thinks is the right answer - slams the brakes on. Kind of a mess.
I doubt any of the anti self-driving car advocates here would accept the common "blinded by the sun" excuse from a driver who had just rear-ended a cyclist. They would unanimously expect that driver to behave like the tesla an those circumstances.
Re: Self driving cars… no thanks.
Might be better to stop gradually, best starting somewhat in advance of the point where visibility becomes restricted (the area covered by fog, the section under the trees, etc) rather than "slamming the brakes on"? It's on the same grounds we have amber between green and red.
Re: Self driving cars… no thanks.
It's down in significant to context and nuance, which is very hard to program and also probably quite unreliable learn from human based training data.Bmblbzzz wrote: ↑13 Jan 2025, 10:26am Might be better to stop gradually, best starting somewhat in advance of the point where visibility becomes restricted (the area covered by fog, the section under the trees, etc) rather than "slamming the brakes on"? It's on the same grounds we have amber between green and red.
Self driving cars in general, unconstrained contexts are very difficult to do right. This isn't news...
Pete.
Often seen riding a bike around Dundee...
Re: Self driving cars… no thanks.
Interestingly of course - the appropriate distance may well be shorter with an autonomous vehicle, because reaction times should be significantly faster.
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.
There are two kinds of people in this world: those can extrapolate from incomplete data.
-
- Posts: 4535
- Joined: 5 Aug 2009, 7:22pm
Re: Self driving cars… no thanks.
Rather worrying if that was the case - because the appropriate distance would also have to include factoring the weather and its effect on road surface/tyre friction. Reaction times are not the be and end all of stopping distances.[XAP]Bob wrote: ↑15 Jan 2025, 9:44amInterestingly of course - the appropriate distance may well be shorter with an autonomous vehicle, because reaction times should be significantly faster.
Dedicated to anyone who has reached that stage https://m.youtube.com/watch?v=Vqbk9cDX0l0 (please note may include humorous swearing)
-
- Posts: 2933
- Joined: 27 Aug 2014, 2:40pm
Re: Self driving cars… no thanks.
What you could do is chain a load of autonomous vehicles together - as they're all together you could then remove the individual engines and just put one big engine up the front to tow them.
It'd be quite efficient to put that whole lot on - oh I dunno , maybe a steel rail or two...?
That way there's no consideration needed for stopping distances between each "car" cos they're all connected in a sort of "train" of cars.
I reckon that could catch on!
Re: Self driving cars… no thanks.
Unlikely as most people drive too close so whilst an AV might have faster reactions, it should also actually follow the need to be able to stop 'within the distance that can be seen to be clear'. This would be in line with any other automated systems (e.g. rail based ones)[XAP]Bob wrote: ↑15 Jan 2025, 9:44amInterestingly of course - the appropriate distance may well be shorter with an autonomous vehicle, because reaction times should be significantly faster.
The contents of this post, unless otherwise stated, are opinions of the author and may actually be complete codswallop
Re: Self driving cars… no thanks.
You actually don't need the rails, or the big engine... if the cars are all talking to each other correctly... then a convoy system is a pretty easy thing to do on motorways.rareposter wrote: ↑16 Jan 2025, 11:31amWhat you could do is chain a load of autonomous vehicles together - as they're all together you could then remove the individual engines and just put one big engine up the front to tow them.
It'd be quite efficient to put that whole lot on - oh I dunno , maybe a steel rail or two...?
That way there's no consideration needed for stopping distances between each "car" cos they're all connected in a sort of "train" of cars.
I reckon that could catch on!
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.
There are two kinds of people in this world: those can extrapolate from incomplete data.
Re: Self driving cars… no thanks.
It could of course also "see" through the car in front by talking to it...Stevek76 wrote: ↑16 Jan 2025, 2:32pmUnlikely as most people drive too close so whilst an AV might have faster reactions, it should also actually follow the need to be able to stop 'within the distance that can be seen to be clear'. This would be in line with any other automated systems (e.g. rail based ones)
There isn't a magic solution here, but you can "look" like you're being aggressive whilst still having reactions fast enough to stop when needed.
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.
There are two kinds of people in this world: those can extrapolate from incomplete data.