omg .. a car that you drive whilst drunk then.. happy days
|
Have I read this report right - the driver still has to be present, and able to take over in an emergency? So a sort of glorified cruise control?
I don't think they are at the stage where you just put in the sat-nav destination and wave it off down the drive?
|
|
Not for me thank you!!! I enjoy driving and have done since I passed my test many years ago. I don't even use cruise control, it just feels weird.
|
>> Have I read this report right - the driver still has to be present, and
>> able to take over in an emergency? So a sort of glorified cruise control?
>>
It's weird, isn't it? The law insists that you're ready to intervene at any time, yet allows you to disconnect from the one thing that might help you remain alert - actually driving the car.
Interesting ethical discussion, too. Suppose the driver of an oncoming car has a heart attack and drifts into your lane. Your car's computer has to make the decision whether to save you by swerving around it - and mowing down the bunch of schoolchildren on the pavement - or by continuing it's course knowing it'll kill you in the process. Somewhere, someone is writing the algorithm to balance the values of human life. Which is a bit unsettling.
|
"Somewhere, someone is writing the algorithm to balance the values of human life. Which is a bit unsettling."
Instead of possibly letting the decision to be made by a 17 year old who has just passed his test and is sending a text message or by an 85 year old with slow reactions and dodgy eyesight. Isn't that just as unsettling ?
Last edited by: CGNorwich on Thu 31 Jul 14 at 09:54
|
>> Instead of possibly letting the decision to be made by a 17 year old who
>> has just passed his test and is sending a text message or by an 85
>> year old with slow reactions and dodgy eyesight. Isn't that just as unsettling ?
>>
I work in IT and some of the errors I have seen in software over the years is incredible.
|
"I work in IT and some of the errors I have seen in software over the years is incredible."
As are some of the errors in driving.
I don't think in practice driverless cars are going to be given unfettered access to all types of roads and conditions from day one . One of the situations is something like a convoy system for vehicles on busy motorways and dual carriageways. You could in theory pack a lot more cars on a stretch of road if they were all being controlled as one.
|
>> One of the situations is
>> something like a convoy system for vehicles on busy motorways and dual carriageways.
Already tested on public roads, in Spain:
www.youtube.com/watch?v=oV3nINN2ELQ
|
That Youtube video was great. If they'd only got Michael Rodd out of retirement to do the narration it would have been exactly like a Tomorrow's World piece from 30 years ago.
Nice change to see old fashioned film making.
As to the content though, not sure about the driver reading. If it does go wrong then I reckon you'd want to be on the ball to take over, not looking at a knitting pattern.
|
>> Instead of possibly letting the decision to be made by a 17 year old who
>> has just passed his test and is sending a text message or by an 85
>> year old with slow reactions and dodgy eyesight. Isn't that just as unsettling ?
>>
>>
Well, that's the thing - if you or me or the 17 year old or the OAP were in that situation we wouldn't be able to react appropriately. We might do nothing, we might swerve instinctively, but ultimately it'd be a manoeuvre that we didn't really have control over. Regardless of the outcome, I think it'd be hard to pin blame and accuse them of taking deliberate action.
The computer, on the other hand, has time to make a calculated decision over what to do. That's what I find unsettling.
|
>>
>> The computer, on the other hand, has time to make a calculated decision over what
>> to do. That's what I find unsettling.
>>
The computer will presumably have to be insured as a named driver.
If the computer is the main driver the policy will be in its name, so would build up a NCD all being well.
It might have to be the registered keeper.
Last edited by: Cliff Pope on Thu 31 Jul 14 at 10:51
|
No need to worry, it is all being sorted. Taken from the BBC website -
"Meanwhile, civil servants have been given until the end of this year to publish a review of road regulations.
This will cover the need for self-drive vehicles to comply with safety and traffic laws, and involve changes to the Highway Code, which applies to England, Scotland and Wales.
Two area will be examined by the review: how the rules should apply to vehicles in which the driver can take back control at short notice, and how they should apply to vehicles in which there is no driver."
>>It might have to be the registered keeper.<< Look forward to the first SP30 ? - the failure to declare the 'driver' will get interesting too.
|
>> >>
>> >> The computer, on the other hand, has time to make a calculated decision over
>> what
>> >> to do. That's what I find unsettling.
>> >>
>>
>> The computer will presumably have to be insured as a named driver.
>> If the computer is the main driver the policy will be in its name, so
>> would build up a NCD all being well.
>> It might have to be the registered keeper.
>>
Ahhh... but what about the person who wrote the software? Shirley his / her actions and decisions about what to write have a direct impact on the behaviour of the vehicle? Who indemnifies them?
It's turtles all the way down... ;-)
|
>>The computer, on the other hand, has time to make a calculated decision over what to do. That's what I find unsettling. <<<
But surely, isn't that what you spend your days in a fly-by-wire cockpit watching? Or are you and the co-pilot asleep? :)
Last edited by: sherlock47 on Thu 31 Jul 14 at 11:12
|
Driverless cars sound great.
If I get a parking ticket, I'll just blame it on the car!
|
>> Driverless cars sound great.
>> If I get a parking ticket, I'll just blame it on the car!
>>
It seems obvious that the "driver" must be capable of reading speed limits and acting accordingly. Presumably it will obey all highway regulations.
Or will it? In many everyday situations it is impossible to make progress if you obey the literal rule of the road. At a busy junction sometimes you just have to give a brief wave and go, regardless of someone else's right of way. Unless the driverless car is just going to sit there all day causing total gridlock, it must be programmed at some point to exercise discretion and break a rule.
Which rules will it be allowed to break? Crucially, who will receive the penalty if it gets caught?
Last edited by: Cliff Pope on Thu 31 Jul 14 at 12:55
|
Cliff, I can envisage a scenario where all these driverless cars are connected to the internet via satellite, and programmes will run to analyse traffic flows at junctions, and send commands to cars to slow down to let others out of side streets for example. Might actually be the end of traffic lights. Evidently only if ALL vehicles are driverless.
What happens during an outage, though, is harder to envisage.
|
>> >>The computer, on the other hand, has time to make a calculated decision over what
>> to do. That's what I find unsettling. <<<
>>
>>
>> But surely, isn't that what you spend your days in a fly-by-wire cockpit watching? Or
>> are you and the co-pilot asleep? :)
>>
Terminology - the autoflight system (that is, the autopilot and flight guidance computers) simply make the job easier. They act on instructions given to them by me (whether that's vertical or horizontal navigation) - they don't have to react quickly and frequently don't. They are closely supervised at all times and they can be disconnected with a single push on a handy red button that sits directly under your thumb. AFAIK, they're roughly analogous to the cruise control features on a car. They're handy, but require careful monitoring and have a tendency to perform less than optimally.
The fly-by-wire system is separate to this entirely and simply has the job of translating instructions given to it by the autopilot (or my sidestick) into pitch attitudes - technically load factors - and roll rates. That's it. They will prevent someone from overstressing, unusual attitudes, overspeed, underspeed, stall etc but much in the same way as you wouldn't say your ABS is driving the car for you, the flight control computers aren't flying the aircraft.
The thing is, do you want your car to hand back control to you when it's heading backwards towards a hedge with a message along the lines, "Sorry guv'nor, it's up to you to fix this...".
|
Naturally I hate the things and don't really believe in them. But they will slowly be imposed in various forms.
God knows there's enough mimsing already. These devices will make it more or less universal. As CP points out, if they can't break the 'law' they will be stymied in many traffic situations. They and the people who fancy them are utterly pathetic.
|
>> God knows there's enough mimsing already. These devices will make it more or less universal.
It could get worse - perhaps they'll be combined with these. 3hp!
tinyurl.com/jwqby6b
EDIT: Well it says 3hp in one place, and 12.3hp somewhere else. Either way, not a lot.
Last edited by: Focusless on Thu 31 Jul 14 at 14:07
|
There was a better one in Australia I seem to remember. Not that much better, but better.
In bright daylight a car whose upper surfaces are electro-whassername can save a bit of fuel. But it really has to be part of a hybrid system.
There was a company that claimed to have devised a sheet metal with some ability to generate power from light, that could be pressed into vehicle bodywork... but I haven't heard or seen anything about it in practical form, and perhaps its efficiency/cost ratio was all wrong (that's what I think anyway). Solar cells have a certain thickness, and a certain frangibility. Pressable sheet solar worth a damn isn't with us yet.
|
>> Naturally I hate the things and don't really believe in them. But they will slowly be imposed in various forms.
Does it seem disgustingly egotistical to remind people of my NLR-published piece 'Prototype Boulevard', on this very subject? I posted a link to it once. But it doesn't say anything new to a proper car person.
|
>> But it doesn't say anything new to a proper car person.
It doesn't say anything you don't know, but it does touch extensively on the areas raised in this thread.
|
>>The thing is, do you want your car to hand back control to you when it's heading backwards towards a hedge with a message along the lines, "Sorry guv'nor, it's up to you to fix this...". <<
Isn't that what happened to AF447?
|
>>
>> >>The thing is, do you want your car to hand back control to you when
>> it's heading backwards towards a hedge with a message along the lines, "Sorry guv'nor, it's
>> up to you to fix this...". <<
>>
>> Isn't that what happened to AF447?
>>
Not really. That was more akin to the car giving you control at 60mph in a straight line, and within ten seconds you somehow manage to get it upside-down on its roof sliding towards a tree, frantically stabbing the button for automatic driving in the hope it'll fix the problem.
The irony of AF447 is that the computer relinquished control of the aircraft in a safe - albeit degraded - mode. Had the pilots sat on their hands for two minutes, in all likelihood they'd have been fine.
Last edited by: Fursty Ferret on Thu 31 Jul 14 at 15:51
|
Didn't know much about AF447, but google came up with this DT feature which seemed pretty informative (would be good if FF could confirm it's basically sound):
www.telegraph.co.uk/technology/9231855/Air-France-Flight-447-Damn-it-were-going-to-crash.html
|
>> Didn't know much about AF447,
>>
There was a documentary on the telly only a couple of weeks which went into the AF447 crash in depth.
The long and short (to a layman) was that the crew were pulling the stick back and putting it into a stall, rather than pushing the stick forward.
Perhaps Fursty could comment/confirm.
|
>> The long and short (to a layman) was that the crew were pulling the stick
>> back and putting it into a stall, rather than pushing the stick forward.
That's what the DT report linked to above says too.
|
Yep, the article is pretty much spot on. Mind you, it is Air France we're discussing here... ;-)
|
>>
>> The long and short (to a layman) was that the crew were pulling the stick
>> back and putting it into a stall, rather than pushing the stick forward.
>>
We've got a fascinating little book in our loo called "Teach Yourself to Fly", written in 1938.
It describes the operation of the joystick, and explains that you will stall if you pull it back too far.
At Chapter 10 it blandly reveals that coming in to land is the most difficult operation - so hastily re-read that chapter while circling for a bit first.
Chapter 11, aerobatics, is a bit more difficult.
In my first, and only, flight in an aeroplane, I was shown the basic operation by my co-pilot, but left the fancy stuff to him.
|
>> We've got a fascinating little book in our loo called "Teach Yourself to Fly",
>> written in 1938.
>> It describes the operation of the joystick, and explains that you will stall if you pull
>> it back too far.
"I taught myself everything I know..." :-)
That's what makes the Airbus great, because (in normal law) pull back on the sidestick as hard as you can with no risk whatsoever of stalling the aircraft. Simply stunning performance.
|
>> That's what makes the Airbus great, because (in normal law) pull back on the sidestick
>> as hard as you can with no risk whatsoever of stalling the aircraft.
So in that very interesting stall video you posted above, have they turned something off to allow them to demonstrate the effects of stalling?
|
>> >> That's what makes the Airbus great, because (in normal law) pull back on the
>> sidestick
>> >> as hard as you can with no risk whatsoever of stalling the aircraft.
>>
>> So in that very interesting stall video you posted above, have they turned something off
>> to allow them to demonstrate the effects of stalling?
>>
Yes, exactly that.
|
Aren't these the cars with women drivers?
(I'll get me coat!!)
|
>> Aren't these the cars with women drivers?
>>
>>
>> (I'll get me coat!!)
>>
i think my misses was reading knitting pattern brochures when she should of (have) been reading the highway code, still absolutely atrocious driving
i hate being in the car with her now
|
If all cars are computerised then there will probably be no need for speed limits on non urban roads as there would be few obstacles for them to negotiate and near instantaneous reaction times.
Imagine motorways being free of tailbacks due to crashes.
I imagine there would be more private tracks for people that wanted a real driving experience.
There could be other advantages.
|
Today's Matt cartoon is funny:
www.telegraph.co.uk/news/matt/
Obviously going to change to something else tomorrow on that link.
|
>> Today's Matt cartoon is funny:
but hopelessly optimistic. When there are enough of them they will turn the whole place into a dreary mimsefest.
|
>> but hopelessly optimistic.
>>
New uses keep coming to mind.
If the car can be sent out on an errand, it could be called up remotely too, eg to fetch someone from the pub or station.
Would need remote phone-operated garage doors and gates too, or perhaps the clever car would see to that rather than just driving through them.
I wonder what it does if it finds itself boxed-in? Bashes someone's bumper and leaves an electronic note perhaps.
|
>> Obviously going to change to something else tomorrow on that link.
If you do prev/next it adds a bit on the end:
www.telegraph.co.uk/news/matt/?cartoon=11008573&cc=11004897
|
That didn't take long.
Not Too Clever User number one:
www.motorauthority.com/news/1093723_this-moron-left-the-drivers-seat-in-a-self-driving-infiniti-q50-on-the-highway-video
This could be fixed by the simple expedient of hooking it into the seat sensor that detects a person, in order to flash up seat belt warnings, but then someone will put an anvil on it.
Last edited by: Crankcase on Thu 7 Aug 14 at 09:26
|
There are times that I really enjoy driving, nothing better than letting a powerful engine roar and propel you along a twisty piece of road.
Yesterday I was travelling to a meeting via the M25 and there was a very nasty accident on the anti-clockwise section between Chertsey and Reigate.
From what I could see, and I certainly did not want to see, a car and a juggernaut had collided the car was totally unrecognisable. I estimate that the car was crushed to 2 or 3 feet. The ambulances weren’t going anywhere.
If driverless cars can stop this sort of thing then they are a good thing.
Of course the option to switch the automation off and enjoy the car is a must!
|
|
Not predict the future but I can see us driving less cars.>:)
|
|
If you are driver who is a passenger in a car you are subconsciously driving anyway. This is almost as mentally tiring as driving, so if you are monitoring an automated car you might just as well drive it. I am sure pilots flying on autopilot get tired, they are monitoring systems and I assume mentally flying the aircraft.
Last edited by: Old Navy on Thu 11 Sep 14 at 07:15
|
Depends who is driving :)
But I wouldn't find it relaxing sitting in the passenger seat of a car with no driver.
|
Not really to do with allowing the driver to have a snooze. Driverless cars would allow a higher density of vehicles on motorways and be more economical in the use of fuel. They would also be potentially safer havering removed the human element which is the major cause of crashes.
|
>> I am sure pilots flying on autopilot get tired, they are monitoring systems and I assume mentally flying the aircraft.
Maybe many do. If the pilot can maintain situational awareness, then at the point when the autopilot disconnects then he or she should be able to intervene and maintain safe flight. But can a pilot really keep up with that for hours on end?
vide AF447. en.wikipedia.org/wiki/Air_France_Flight_447
I'm not a nervous flyer, but that really shook me.
Driverless cars will come of course, and when they do the vast majority of passengers will let them get on with it, and just read their papers, watch TV, or play with their phones.
|
driver -less cars? - Cliff Pope
