Pebble wrote:Having recently acquired a state of the art high tech car with all manner of driver aids, I'm far from impressed, the autonomous anti collision braking system goes off unexpectedly for no reason, the autonomous parking feature can't find anywhere to park in a near empty car park, and even the speed sign recognition feature fails more often than not, the new 20mph limit has it totally bamboozled.
What happens when the automatic braking activates... does it bring the car to a halt? Does it tell you why it thinks it did it?
I don't know that there was a production car sold in the UK that was claimed to be able to find a parking place... what is it, please?
People can dissect a self driving car video as much as they like and play "what if'" and "what about" but they'd be missing an important point.
So far self driving cars have clocked up 100's of millions of real world miles between them (and tens of billions in simulations) and they've proven to be an order of magnitude safer than apes.
Are they perfect? Of course not and there are undoubtedly many scenarios they can't handle particularly well yet but they're built on data and the more data they have the better they get ("experience" if you like). Tesla have recently enabled a fairly high level of automation on their cars, they'll be sucking up terabytes of data from their cars now and every hiccup will be being used to improve their systems, in their labs the simulations will be replaying those hiccups millions of times over and over until they get them right. From what I can gather updates are coming out pretty much weekly at the moment.
Ten years ago we barely had cars that could steer in lane, today you can buy an off the shelf car that will drive you to work and back with minimal intervention, where will we be in another 10?
Although my recent car experience has been more positive than pebble's, (I have a new-ish so-called mini) I have some sympathy in that the computer-controlled elements of the vehicle almost get in the way of driving. My previous car, an excellent little diesel Lupo, had no electronic driver's aids whatsoever (other than a minute lcd-disply clock...) and was much more staightforward to drive. The current jobbie is forever beeping at me to remind me of some obscure issue or other, most of which passes straight over my head. Still have to find my own parking spaces, though...
S
(on the look out for Armageddon, on board a Brompton nano & ever-changing Moultons)
I do wonder if, when self-driving cars become a thing on our streets, some human drivers will take risks because they see them as being over cautious - in much the same way as they would with cyclists and learner drivers, for example; hasty overtaking, driving too close.
The older I get the more I’m inclined to act my shoe size, not my age.
peetee wrote:I do wonder if, when self-driving cars become a thing on our streets, some human drivers will take risks because they see them as being over cautious - in much the same way as they would with cyclists and learner drivers, for example; hasty overtaking, driving too close.
They already do - where over cautious means driving at or near to the speed limit. Loads of great examples of such behaviour on YouTube.
Next big gov announcement: from 2040 all new cars must be self driving with dual mode vehicles banned by 2045. (TBH I suspect insurance costs of DIY driving will be the limiting factor in the future - that and the need for a license).
pwa wrote:In busy traffic the driverless car would have stayed behind the cyclist, unable to get out into the passing lane because of an endless stream of traffic in that lane and the unwillingness of the control system to pull out at the first sign of a gap, the way a confident driver might. Manual over-ride would have been employed. But which is right? Being stuck but safe behind the cyclist for five minutes, or being more pushy and assertive and edging out into the flow of traffic in the adjacent lane? It would be better if the control systems of different vehicles interacted so that one vehicle would slow to let another out.
You mean when this tech becomes endemic the AI starts cooperating to mutual benefit, hmmm, you might be on to something there. I like your window
I think Tesla and others will probably push on with developing the technology
In the video the driver changes the view on the screen as the situation develops and before the car does anything. (That my small phone screen shows) And after the cyclists are passed he changes back to "full road ahead view". This suggests that either the situation is pre- planned or that the driver saw and anticipated BEFORE the auto system did. Which suggests something?
As for all the auto cars communicating with each other, in some sort of optimal road "ballet", I remain hugely sceptical about the ability of even multiple sensors on every vehicle to deliver such a complexity with anything even approaching acceptability. And certainly nothing as complex as that which human drivers on busy motorways achieve every rush hour daily in all weather conditions.
Not to mention a 100% reliability of inter vehicle communication over hundreds of vehicles across thousands of square meters.
IMV fantasy land. A human settlement on Mars is more feasible! And less likely, or desirable.
Lots of things we take for granted were once fantasy.
peetee wrote:I do wonder if, when self-driving cars become a thing on our streets, some human drivers will take risks because they see them as being over cautious - in much the same way as they would with cyclists and learner drivers, for example; hasty overtaking, driving too close.
They already do - where over cautious means driving at or near to the speed limit. Loads of great examples of such behaviour on YouTube.
Next big gov announcement: from 2040 all new cars must be self driving with dual mode vehicles banned by 2045. (TBH I suspect insurance costs of DIY driving will be the limiting factor in the future - that and the need for a license).
DIY driving to be confined to designated tracks etc where the humans can happily collide to their hearts' content.
I think it's: 1. Impatience 2. A desire for excitement, danger, freedom etc.
This appetite for risk varies a lot between people, but I think it is fair to say that it is usually strongest in young men. There are various outlets for this urge for risk, and some are much better than others. I think mountain biking is one of the best, and fast driving definitely one of the worse ones.
If we seek to squash down the opportunities for risk-taking in one area of life, will it pop up somewhere else, perhaps in a worse form?
I think it's: 1. Impatience 2. A desire for excitement, danger, freedom etc.
This appetite for risk varies a lot between people, but I think it is fair to say that it is usually strongest in young men. There are various outlets for this urge for risk, and some are much better than others. I think mountain biking is one of the best, and fast driving definitely one of the worse ones.
If we seek to squash down the opportunities for risk-taking in one area of life, will it pop up somewhere else, perhaps in a worse form?
It's a fair observation I think. Statistically, young men are by far the most likely to participate in accidents, I believe.
StephenW wrote:Why do people drive quickly? I think it's: 1. Impatience 2. A desire for excitement, danger, freedom etc.
I'd include the belief that they can get where they want to be sooner by doing so. And pleasure/ fun in 2
StephenW wrote:If we seek to squash down the opportunities for risk-taking in one area of life, will it pop up somewhere else, perhaps in a worse form?
Not necessarily.
NB the repeatedly expressed desire by a few people in this forum to stop other people doing things. We should always be clear why. The risk to others on public roads is a pretty clear example for why freedom might be appropriately restricted.
