Why driverless car's are a LONG way off.
Discussion
sebhaque said:
Hi pal 
I read an article related to your terrorist theory about Ford's OnStar system (in the USA). Can't find the article in question but this one is pretty similar in terms of results:
http://www.forbes.com/sites/andygreenberg/2013/07/...
It's gone midnight so I'll save writing my thoughts for a more decent hour in the day, but I do see a benefit of driverless cars. While they're somewhat imperfect technology at the moment, the multitude of driver aids in cars like the Tesla particularly, are bound to be a stepping stone towards much more autonomous cars.
Can I see cars being allowed to be fully autonomous? No. Can I see fully autonomous cars with a driver simply there to take over if something goes wrong? Yes. I would imagine within the next couple of decades that most cars will be able to handle themselves for 90% of driving situations, but will require someone behind a wheel of sorts ready to take over should the car experience a dilemma or for any type of failure. Humans accept other humans crashing into things, yet even if there's one autonomous accident in the collective worldwide spectrum in a single year, that'll be all over the media.
Hello mate!I read an article related to your terrorist theory about Ford's OnStar system (in the USA). Can't find the article in question but this one is pretty similar in terms of results:
http://www.forbes.com/sites/andygreenberg/2013/07/...
It's gone midnight so I'll save writing my thoughts for a more decent hour in the day, but I do see a benefit of driverless cars. While they're somewhat imperfect technology at the moment, the multitude of driver aids in cars like the Tesla particularly, are bound to be a stepping stone towards much more autonomous cars.
Can I see cars being allowed to be fully autonomous? No. Can I see fully autonomous cars with a driver simply there to take over if something goes wrong? Yes. I would imagine within the next couple of decades that most cars will be able to handle themselves for 90% of driving situations, but will require someone behind a wheel of sorts ready to take over should the car experience a dilemma or for any type of failure. Humans accept other humans crashing into things, yet even if there's one autonomous accident in the collective worldwide spectrum in a single year, that'll be all over the media.
I totally agree with you. I'm actually looking forward to autonomous motorway driving, it is a chore most of the time and traffic jams caused by drivers following too closely will be a thing of the past.
I do think there are too many scenarios and potential hazards on back roads and in built up areas for autonomous cars to take over completely.
Also worth noting it will take a long time for less developed countries to follow suit.
kambites said:
skyrover said:
Tell me... what would autopilot do with no pilot, untrustworthy airspeed data, a hydraulic leak and potential engine failure?
Conversely, there have been commercial airliner crashes caused by to pilots not believing perfectly functional instruments and, for example, shutting down the wrong engine when one has failed. Commercial aeroplane auto-pilots aren't really designed to operate with no human intervention. If they were, I suspect they could deal with such problems at least as well as human pilots.
skyrover said:
Can you tell me how you would design a computer to operate after it has turned itself off due to short circuit/fire?
Can you tell me how a pilot will fly a completely FBW airliner after he/it has turned off the computers due to a short circuit/fire? Using the back-up computers, one would assume. These problems are real, but they're not an issue with automation; they're an issue with electronic control which is a completely different thing. Some modern air-liners cannot be flown, either by the pilot or auto-pilot, without computers and possibly a majority can't be flown without electricity.
Edited by kambites on Wednesday 1st June 19:44
skyrover said:
This is true... however I would trust the decision making of a well trained pilot over a computer.
That's fair enough; there's obviously no right or wrong in these things but personally I'm the other way around even with commercial air-line pilots and I trust commercial air-line pilots an awful lot more than private car drivers. kambites said:
I think if truly driver-less cars do happen, they'll dramatically change the way we think about cars.
They will fundamentally change the way we think about travel full stop.As it stands most us us own or 'own' a car. For most people it will be not in use maybe 22 hours in every 24. And when it is in use it only carries 1 out of 4 possible passengers. That's a massive, criminal waste of capacity.
Pool cars of some description have the potential to make use of that capacity much more efficiently, spreading the load and reducing congestion, emissions, cost, space needed for parking, the list goes on. That is if you can get away from the idea that a car is your own personal fiefdom to charge around in for a hour a day, but to leave rotting in a car park the rest of the time.
In here I'll not hold my breath. Fortunately those who matter are more enlightened.
technodup said:
hey will fundamentally change the way we think about travel full stop.
As it stands most us us own or 'own' a car. For most people it will be not in use maybe 22 hours in every 24. And when it is in use it only carries 1 out of 4 possible passengers. That's a massive, criminal waste of capacity.
Pool cars of some description have the potential to make use of that capacity much more efficiently, spreading the load and reducing congestion, emissions, cost, space needed for parking, the list goes on. That is if you can get away from the idea that a car is your own personal fiefdom to charge around in for a hour a day, but to leave rotting in a car park the rest of the time.
In here I'll not hold my breath. Fortunately those who matter are more enlightened.
Then what would happen if you needed to go to work but the pool cars were all taken, or wouldn't get here for 2 hours? Or you need to make an impromptu/emergency trip?As it stands most us us own or 'own' a car. For most people it will be not in use maybe 22 hours in every 24. And when it is in use it only carries 1 out of 4 possible passengers. That's a massive, criminal waste of capacity.
Pool cars of some description have the potential to make use of that capacity much more efficiently, spreading the load and reducing congestion, emissions, cost, space needed for parking, the list goes on. That is if you can get away from the idea that a car is your own personal fiefdom to charge around in for a hour a day, but to leave rotting in a car park the rest of the time.
In here I'll not hold my breath. Fortunately those who matter are more enlightened.
Mr Snrub said:
Then what would happen if you needed to go to work but the pool cars were all taken, or wouldn't get here for 2 hours? Or you need to make an impromptu/emergency trip?
If it becomes a significant proportion of the market that just wont happen; as the number of vehicles rises the odds of an anomaly in terms of availability falls. I'm sure some people will continue to own cars for the foreseeable future, probably most people on here among them. I suppose initially we'll start to see two cars in one household become far rarer. Even if the technology was available cheaply and reliably enough now, things wouldn't change over-night; society is very resistant to sudden change.
If we do get to that point, it would certainly change what I own. I bought my car as a compromise between usability for commuting and fun; if I could simply call up a taxi to take me to and from work for 50p, I'd be running a proper toy instead.
Edited by kambites on Wednesday 1st June 20:15
Mr Snrub said:
technodup said:
hey will fundamentally change the way we think about travel full stop.
As it stands most us us own or 'own' a car. For most people it will be not in use maybe 22 hours in every 24. And when it is in use it only carries 1 out of 4 possible passengers. That's a massive, criminal waste of capacity.
Pool cars of some description have the potential to make use of that capacity much more efficiently, spreading the load and reducing congestion, emissions, cost, space needed for parking, the list goes on. That is if you can get away from the idea that a car is your own personal fiefdom to charge around in for a hour a day, but to leave rotting in a car park the rest of the time.
In here I'll not hold my breath. Fortunately those who matter are more enlightened.
Then what would happen if you needed to go to work but the pool cars were all taken, or wouldn't get here for 2 hours? Or you need to make an impromptu/emergency trip?As it stands most us us own or 'own' a car. For most people it will be not in use maybe 22 hours in every 24. And when it is in use it only carries 1 out of 4 possible passengers. That's a massive, criminal waste of capacity.
Pool cars of some description have the potential to make use of that capacity much more efficiently, spreading the load and reducing congestion, emissions, cost, space needed for parking, the list goes on. That is if you can get away from the idea that a car is your own personal fiefdom to charge around in for a hour a day, but to leave rotting in a car park the rest of the time.
In here I'll not hold my breath. Fortunately those who matter are more enlightened.
Mr Snrub said:
Then what would happen if you needed to go to work but the pool cars were all taken, or wouldn't get here for 2 hours? Or you need to make an impromptu/emergency trip?
Your hypothetical problem situation doesn't stand up to the savings I listed.Bottom line is there would be availability. Uber already do something similar with surge pricing, supply and demand- it's really not that complex. Of course in the country times would be longer than the city but that's the case now for driving, trains, deliveries etc.
If you think an individual's unlikely but statistically possible inconvenience is going to trump fewer road deaths, lower emissions, huge profits, lower costs yadda yadda I think you'll be disappointed.
skyrover said:
I've seen plenty of posts recently recently describing the imminent demise of the human driver in place of autonomous car's and how our leadership will soon look to ban human piloted car's altogether on the grounds of "safety".
I am going to list the reasons why this simply is not the case, at least for the near and somewhat distant foreseeable future.
1. Humans are actually very good at driving with around 1 death for every 100 million miles driven on average, something a machine will find incredibly difficult to match for the following reasons.
2. Machines are incapable of dealing with tasks that have not been foreseen by the software engineers and each possible eventuality programmed with acceptable solutions. It will take billons of miles driven and an absolutely enormous variety of situations and hazards for the software engineers understand and to solve.
3. Machines are incapable of making philosophical decisions. If a crash is unavoidable, who do you hit, the lady crossing the street or the oncoming car? How does a machine anticipate hazards i.e is the child on a bicycle along the pavement more likely to veer into the road compared to the adult?
4. How does a machine react to faulty sensors.. or sensors giving inaccurate information?
5. Someone is about to drive into you... do you continue on or stop for the red light in front of you?
6. How does the machine behave with damage or neglect? etc etc
Even google.. arguably the most successful implementer of driverless technology so far admits it's going to be a long time before these vehicles are a common sight on our roads.
http://thenextweb.com/opinion/2016/03/18/great-tec...
Then there are legal, legislative and consumer acceptance barriers to get through, not to mention inter vehicle software communication/update standards to decide upon.
So sleep easy folks, ignore the hype, your steering wheel and pedals are safe for a long while yet.
I know I'm wasting my time already (also, if no one else has pointed it out, it should be "cars" not "car's" in the title, which is never going to put me in an optimistic frame of mind), but some brief notes:I am going to list the reasons why this simply is not the case, at least for the near and somewhat distant foreseeable future.
1. Humans are actually very good at driving with around 1 death for every 100 million miles driven on average, something a machine will find incredibly difficult to match for the following reasons.
2. Machines are incapable of dealing with tasks that have not been foreseen by the software engineers and each possible eventuality programmed with acceptable solutions. It will take billons of miles driven and an absolutely enormous variety of situations and hazards for the software engineers understand and to solve.
3. Machines are incapable of making philosophical decisions. If a crash is unavoidable, who do you hit, the lady crossing the street or the oncoming car? How does a machine anticipate hazards i.e is the child on a bicycle along the pavement more likely to veer into the road compared to the adult?
4. How does a machine react to faulty sensors.. or sensors giving inaccurate information?
5. Someone is about to drive into you... do you continue on or stop for the red light in front of you?
6. How does the machine behave with damage or neglect? etc etc
Even google.. arguably the most successful implementer of driverless technology so far admits it's going to be a long time before these vehicles are a common sight on our roads.
http://thenextweb.com/opinion/2016/03/18/great-tec...
Then there are legal, legislative and consumer acceptance barriers to get through, not to mention inter vehicle software communication/update standards to decide upon.
So sleep easy folks, ignore the hype, your steering wheel and pedals are safe for a long while yet.
1) on safety: as anecdote, I work at a chemical plant with enough heavy machinery to destroy several terminators, and, well, let's just say you really don't want to get any of this stuff on your shoes. It's an aging, complex plant at that. Over the last decade, excluding heart attacks, 100% of the major health incidents have been from people driving, in daylight hours. (and that 100% is a statistically significant number). The nearest competing site has a much better safety record with one major distinction: no private vehicles on site. This does not include incidents involving workers driving to and from site (or other work business). Overall, the entirety of work fatalities in the country is about a tenth of the road fatalities, and we spend a lot more time at work.
By any modern work standard except medicine, where stabbing and poisoning people is sort of necessary, driving is the most dangerous thing we do, and worse, a huge amount of that risk is borne by pedestrians, cyclists, other motorists - people who never had the choice to let you drive or not, but still have to bear the consequences.
3) the world would be a better place if anyone who considered themselves an ethical driver, especially in a pinch, was taken out and shot. For a start, in an emergency, you are NOT going to be thinking that deeply. Most likely you'll choose to avoid the biggest thing in your central vision, because that's how your caveman great-grandaddy survived best. Second, the people doing 68 in the outside lane because "that's fast enough"? "Ethical" driver. The one accelerating and flashing to block an overtake, or zip merge? "Ethical" driver. The absolute last thing the roads need is philosophers. Want to drive well? Keep a look out, don't go faster than you can stop, follow predictable rules.
If you do that and someone still dies, the best thing to do would be to lock up the meatbag responsible. Unless they're the one that died, in which case problem solved. If they're a small meatbag things are slightly different, but honestly not that much. You as a driver can still only be asked to do the same things.
If it was the set of rules/logic at fault (small chance, but not zero), then even better - so long as those who sell (and regulate the sale of) these cars have even the tiniest bit of sense (and funny as the inevitable joke is, they probably do have a bit), that problem will be banished from the firmware within a couple of weeks. Maybe a few months for a particularly natty problem, and that WOULD be a disaster, since it would likely necessitate workarounds that will probably cripple performance.
For every car.
Can you imagine that? No fatal mistake made by a machine on the roads will ever need to be made again.
The only real problem is managing the changeover, when the robo-cars may be overly cautious, and dealing with owners who may have never really grasped the responsibility involved. Lord knows I can barely hold a piece of electronics in my hand for more than twenty minutes without trying to install linux on it, and you're asking me, or an impatient powerfully-built director, to just let the car trundle through an area at 20 for no obvious reason? hmm.
quote=kambites]
Can you tell me how a pilot will fly a completely FBW airliner after he/it has turned off the computers due to a short circuit/fire? Using the back-up computers, one would assume.
These problems are real, but they're not an issue with automation; they're an issue with electronic control which is a completely different thing. Some modern air-liners cannot be flown, either by the pilot or auto-pilot, without computers and possibly a majority can't be flown without electricity.
[/quote]
On at least two occasions, commercial airliners have been landed with no control surface capability at all. Good (somewhat understated) presentation here:
http://www.smartcockpit.com/download.php?path=docs...
After 70 years of autopilot development, I would not go anywhere near a pilotless aircraft - in theory it can work, they can take off automatically, they can fly, they can land automatically. However, as soon as something goes wrong, you need the human. Sometimes the human gets it wrong - on AF447, they certainly did, but the autopilot gave up first because it could not understand the data it was receiving. And to be clear, flying is a much easier problem to solve than driving - you've got a small army of people ensuring that you don't get too close to the 'plane in front.
This will inevitably happen, it will just take a lot longer than most people expect. Nor do I see fundamental changes in car ownership patterns - humans are not logical. If we were logical, we'd all be driving Kias, which would get you to work just as efficiently as a BMW or any other car that we cheerfully pay for today.
Can you tell me how a pilot will fly a completely FBW airliner after he/it has turned off the computers due to a short circuit/fire? Using the back-up computers, one would assume.
These problems are real, but they're not an issue with automation; they're an issue with electronic control which is a completely different thing. Some modern air-liners cannot be flown, either by the pilot or auto-pilot, without computers and possibly a majority can't be flown without electricity.
Edited by kambites on Wednesday 1st June 19:44
[/quote]
On at least two occasions, commercial airliners have been landed with no control surface capability at all. Good (somewhat understated) presentation here:
http://www.smartcockpit.com/download.php?path=docs...
After 70 years of autopilot development, I would not go anywhere near a pilotless aircraft - in theory it can work, they can take off automatically, they can fly, they can land automatically. However, as soon as something goes wrong, you need the human. Sometimes the human gets it wrong - on AF447, they certainly did, but the autopilot gave up first because it could not understand the data it was receiving. And to be clear, flying is a much easier problem to solve than driving - you've got a small army of people ensuring that you don't get too close to the 'plane in front.
This will inevitably happen, it will just take a lot longer than most people expect. Nor do I see fundamental changes in car ownership patterns - humans are not logical. If we were logical, we'd all be driving Kias, which would get you to work just as efficiently as a BMW or any other car that we cheerfully pay for today.
technodup said:
I imagine the public are happier with the boffins finding the solutions than the PH flat earthers finding the 'problems'. I know I am.
Me too. After 6 years of work, and thousands of miles driven autonomously, I can't imagine Ian Google is going to read the OP's post and say 'f
k me, we never thought of any of that, tools down lads, this'll never work'. You can go out and buy a Tesla that will happily pilot itself down the motorway TODAY, it's a bit odd to think driverless cars are 'a long way off'.
I wasn't attempting to say that current autopilot systems fitted to commercial planes could operate in place of pilots; that's not what they're designed to do so as you say they lack redundancy and the ability to deal with faults. I was saying that with current technology it would be perfectly possible to build an auto-pilot system which I would trust more than a pilot overall.
I think the pool car idea is doomed to failure due to human nature undortunately, the only reason taxis aren't overflowing out the sunroof with piss, st, vomit and fagends is the driver is there to regulate the behaviour of the large portion of society that behave like animals when there isn't someone stopping them.
t, vomit and fagends is the driver is there to regulate the behaviour of the large portion of society that behave like animals when there isn't someone stopping them.rxe said:
1. Driving circumstances, not get it right 100% of the time.
2. Simple example. There's an accident on the road ahead, plod have got there and there is a temporary diversion. Mr plod is directing you off the road onto a track that connects with the road 200 yards ahead. A human understands this, and does what he's told. At the moment, a driverless car could not understand this. Or how about freshly fallen snow? All the signs have gone, the road has gone. Can it still cope? How about you ask your car to pick you up from some event. It cannot comprehend that it needs to park in a field as directed by a marshall - which leads to a massive queue of cars as they all fail to cope with a simple instruction.
3. As to sensors - 'planes are frequently grounded by single failed sensors. They will have redundancy, but they still don't take off. And another point on aviation - we've had autopilot since WWII, 70 years later, no one is seriously talking about getting rid of very expensive pilots.
1. Drivers fail to handle the circumstances, and crash as a result.2. Simple example. There's an accident on the road ahead, plod have got there and there is a temporary diversion. Mr plod is directing you off the road onto a track that connects with the road 200 yards ahead. A human understands this, and does what he's told. At the moment, a driverless car could not understand this. Or how about freshly fallen snow? All the signs have gone, the road has gone. Can it still cope? How about you ask your car to pick you up from some event. It cannot comprehend that it needs to park in a field as directed by a marshall - which leads to a massive queue of cars as they all fail to cope with a simple instruction.
3. As to sensors - 'planes are frequently grounded by single failed sensors. They will have redundancy, but they still don't take off. And another point on aviation - we've had autopilot since WWII, 70 years later, no one is seriously talking about getting rid of very expensive pilots.
2. In my 40 odd years of driving I don't think I've ever encountered that. (I've driven on tracks by mistake though). But you might be missing a massive point. The reason that roads have to be closed are often because drivers can not be trusted to behave properly at an accident scene - they might speed, or shout at the emergency workers, or film the scene as they drive through. Driverless cars could be trusted to pass through the scene safely.
In any case the odds of a convenient track being in place are so remote that we can just forget that.
Here's an example for you though - true story. Truck breaks down on a dc in conditions of perfect visibility. Two cars approach but both drivers fail to give way to each other and attempt to pass side-by-side. They can't so they stop. Motorcyclist comes up behind at speed and crashes into the scene.
3. As you know, pilotless drones are striking targets with phenomenal accuracy (irrespective of whether the target is correct or not).
However we should be careful about comparisons with aircraft, because when they stop working they can fall out od the sky. Cars won't do that. When a car stops unexpectedly it can cause drivers enormous problems but I don't think this give driverless cars any - I imagine the driverless cars are communicating each other so will know a car has stopped ahead.
Meantime, drivers running into the car in front is the most common type of collision.
OldGermanHeaps said:
I think the pool car idea is doomed to failure due to human nature undortunately, the only reason taxis aren't overflowing out the sunroof with piss, st, vomit and fagends is the driver is there to regulate the behaviour of the large portion of society that behave like animals when there isn't someone stopping them.
Not only that, but we already have hugely congested roads in a lot of the country. If you also add empty vehicles cluttering up the roads on the way to pick people up then it could potentially be like an all day rush hour.t, vomit and fagends is the driver is there to regulate the behaviour of the large portion of society that behave like animals when there isn't someone stopping them.kambites said:
I'd like to see your evidence for this rather enormous leap of logic?
Our politicians like to ban things that people enjoy for..... reasons.....You really think we aren't next in line after they've finished with smokers?
Elon Musk thinks so
Gassing Station | General Gassing | Top of Page | What's New | My Stuff