Driverless cars and the ownerless future
Discussion
I've always liked the idea of autonomy, who wouldn't want to be driven to the pub and back, but being a Londoner I have never believed it to be possible to achieve in the timeframes, or even remotely close, being suggested. Systems may be able to navigate highly simplistic road sections 99% of the time but when you look outside of things like the grid system you soon see that the tech is nowhere near any level of true competency and that there is one long term, terminal issue facing autonomous cars across the planet with the exception of Vancouver. And that is the pedestrian.
The fundamental flaw with autonomy is that as a pedestrian, once you know that the large, hurtling, leathal metal box is fully programmed to halt before your magnificence then those cars will never move in the blended urban environment. The only reason that pedestrians or cyclists give way to cars is because the car represents a big safety variable and therefor risk. Once the car is programmed to be reliably subservient then man is released from that fear to take advantage of this new freedom.
In shabbier parts of London you have always had the local tough guy who will select a vulnerable looking road user, make eye contact with the driver and step out into the road, forcing the driver to stop for them. Apply alcohol and the number of men who reveal themselves as that sort of person climbs steeply.
In places like Oxford St with high pedestrian footfall, when trying to cross from North to South or vice verse the traffic lights are largely irrelevant as the pedestrians keep crossing until the lead car specifically moves forward at them. It is only that movement that triggers an alert to pay attention to the rules. You already see that when the lead car is a tourist, not a resident, they end up never crossing until enough cars behind are blowing their horns. Cars programmed to not specifically drive at pedestrians will just sit there all day.
And the City? Home of the alpha male and wannabe alpha male. Once cars are programmed to give way then they will never move again.
As for any other part of London, I've already got used to Tesla's coming to a halt in the road because a pedestrian is waiting between two parked cars to cross and has, as always, positioned themselves forward of the line of parked cars to help vision and to shorten the crossing space.
Even on motorways which are arguably the simplest roads in the UK you can see the current tech isn't yet ready to cope with the huge array of curveballs and the dynamic environment that even the least powerful human brain can handle really quite easily.
https://www.reuters.com/article/us-uber-atg/uber-s...
The news that Uber is offloading it's autonomous division from its balance sheet just confirms how far away we are and remain. They've taken it off their books because it is not going to deliver revenue within any foreseeable economic timeframe. And in order to do that they've had to create a new entity to 'buy' it from them and to spread the ongoing capital drain.
Autonomy has had a hugely US centric mindset which has enabled it to attract vast funds but the reality is that the US grid system, absence of pedestrians and cyclists and lack of jay walkers is not remotely representative of the world, it is in reality highly anomalous. Even on such young and simplified networks with very limited variance in road users they aren't getting autonomy to work with vehicles hitting lorries, static cars, walls and obviously that poor sod on a bike.
The real tests and development are the programs being runnin places like London and Oxford, old, complex systems with a huge array of different forms of transport all blended in mixed together. Even Oxford is far more formulaic and reliable than London as the locals whthwrnon foot, bike or car are more law obiding and less self centred.
It's the cracking of old metropolis such as London that will signify the viability of autonomy and I don't think that's happening anytime soon.
The fundamental flaw with autonomy is that as a pedestrian, once you know that the large, hurtling, leathal metal box is fully programmed to halt before your magnificence then those cars will never move in the blended urban environment. The only reason that pedestrians or cyclists give way to cars is because the car represents a big safety variable and therefor risk. Once the car is programmed to be reliably subservient then man is released from that fear to take advantage of this new freedom.
In shabbier parts of London you have always had the local tough guy who will select a vulnerable looking road user, make eye contact with the driver and step out into the road, forcing the driver to stop for them. Apply alcohol and the number of men who reveal themselves as that sort of person climbs steeply.
In places like Oxford St with high pedestrian footfall, when trying to cross from North to South or vice verse the traffic lights are largely irrelevant as the pedestrians keep crossing until the lead car specifically moves forward at them. It is only that movement that triggers an alert to pay attention to the rules. You already see that when the lead car is a tourist, not a resident, they end up never crossing until enough cars behind are blowing their horns. Cars programmed to not specifically drive at pedestrians will just sit there all day.
And the City? Home of the alpha male and wannabe alpha male. Once cars are programmed to give way then they will never move again.
As for any other part of London, I've already got used to Tesla's coming to a halt in the road because a pedestrian is waiting between two parked cars to cross and has, as always, positioned themselves forward of the line of parked cars to help vision and to shorten the crossing space.
Even on motorways which are arguably the simplest roads in the UK you can see the current tech isn't yet ready to cope with the huge array of curveballs and the dynamic environment that even the least powerful human brain can handle really quite easily.
https://www.reuters.com/article/us-uber-atg/uber-s...
The news that Uber is offloading it's autonomous division from its balance sheet just confirms how far away we are and remain. They've taken it off their books because it is not going to deliver revenue within any foreseeable economic timeframe. And in order to do that they've had to create a new entity to 'buy' it from them and to spread the ongoing capital drain.
Autonomy has had a hugely US centric mindset which has enabled it to attract vast funds but the reality is that the US grid system, absence of pedestrians and cyclists and lack of jay walkers is not remotely representative of the world, it is in reality highly anomalous. Even on such young and simplified networks with very limited variance in road users they aren't getting autonomy to work with vehicles hitting lorries, static cars, walls and obviously that poor sod on a bike.
The real tests and development are the programs being runnin places like London and Oxford, old, complex systems with a huge array of different forms of transport all blended in mixed together. Even Oxford is far more formulaic and reliable than London as the locals whthwrnon foot, bike or car are more law obiding and less self centred.
It's the cracking of old metropolis such as London that will signify the viability of autonomy and I don't think that's happening anytime soon.
DonkeyApple said:
I've always liked the idea of autonomy, who wouldn't want to be driven to the pub and back, but being a Londoner I have never believed it to be possible to achieve in the timeframes, or even remotely close, being suggested.
The vehicles might be autonomous but if you have to wait for one to turn up, it limits your own autonomy. If you ran a fleet of autonomous cars (driver-less taxis) would you have them as utilised as possible doing paid work or have them hanging around waiting for a fare?Cleanliness is one issue with the ownerless model that is rarely mentioned.
I worked as a private hire driver whilst at uni and if there's one lesson I learnt, it's that human beings are disgusting! Add alcohol to the mix and they become really disgusting!
I had quite a nice A6 to drive around and did my best to keep it clean but it was a never ending thankless task. Between almost every job I was disposing of rubbish, every 3 or 4 jobs I'd be picking up dropped chips and wiping up sauce from spilled kebabs. That's all before we get to the bodily fluids. The car then needed properly cleaned, hoovered and wiped down after every shift.
Now, if they behave like that with the owner of the car sitting next to them asking them not to eat, how are they going to behave if they're on their own?
I worked as a private hire driver whilst at uni and if there's one lesson I learnt, it's that human beings are disgusting! Add alcohol to the mix and they become really disgusting!
I had quite a nice A6 to drive around and did my best to keep it clean but it was a never ending thankless task. Between almost every job I was disposing of rubbish, every 3 or 4 jobs I'd be picking up dropped chips and wiping up sauce from spilled kebabs. That's all before we get to the bodily fluids. The car then needed properly cleaned, hoovered and wiped down after every shift.
Now, if they behave like that with the owner of the car sitting next to them asking them not to eat, how are they going to behave if they're on their own?
The 'ownerless' model won't work. As soon as there is a lack of availability, the whole thing falls over.
Which of course there will be, because journeys are irregular. Not everyone simply commutes from A to B, every day, at the same time. In fact that model is massively in decline, post-Covid.
Which of course there will be, because journeys are irregular. Not everyone simply commutes from A to B, every day, at the same time. In fact that model is massively in decline, post-Covid.
DonkeyApple said:
I've always liked the idea of autonomy, who wouldn't want to be driven to the pub and back, but being a Londoner I have never believed it to be possible to achieve in the timeframes, or even remotely close, being suggested. Systems may be able to navigate highly simplistic road sections 99% of the time but when you look outside of things like the grid system you soon see that the tech is nowhere near any level of true competency and that there is one long term, terminal issue facing autonomous cars across the planet with the exception of Vancouver. And that is the pedestrian.
The fundamental flaw with autonomy is that as a pedestrian, once you know that the large, hurtling, leathal metal box is fully programmed to halt before your magnificence then those cars will never move in the blended urban environment. The only reason that pedestrians or cyclists give way to cars is because the car represents a big safety variable and therefor risk. Once the car is programmed to be reliably subservient then man is released from that fear to take advantage of this new freedom.
In shabbier parts of London you have always had the local tough guy who will select a vulnerable looking road user, make eye contact with the driver and step out into the road, forcing the driver to stop for them. Apply alcohol and the number of men who reveal themselves as that sort of person climbs steeply.
In places like Oxford St with high pedestrian footfall, when trying to cross from North to South or vice verse the traffic lights are largely irrelevant as the pedestrians keep crossing until the lead car specifically moves forward at them. It is only that movement that triggers an alert to pay attention to the rules. You already see that when the lead car is a tourist, not a resident, they end up never crossing until enough cars behind are blowing their horns. Cars programmed to not specifically drive at pedestrians will just sit there all day.
And the City? Home of the alpha male and wannabe alpha male. Once cars are programmed to give way then they will never move again.
As for any other part of London, I've already got used to Tesla's coming to a halt in the road because a pedestrian is waiting between two parked cars to cross and has, as always, positioned themselves forward of the line of parked cars to help vision and to shorten the crossing space.
Even on motorways which are arguably the simplest roads in the UK you can see the current tech isn't yet ready to cope with the huge array of curveballs and the dynamic environment that even the least powerful human brain can handle really quite easily.
https://www.reuters.com/article/us-uber-atg/uber-s...
The news that Uber is offloading it's autonomous division from its balance sheet just confirms how far away we are and remain. They've taken it off their books because it is not going to deliver revenue within any foreseeable economic timeframe. And in order to do that they've had to create a new entity to 'buy' it from them and to spread the ongoing capital drain.
Autonomy has had a hugely US centric mindset which has enabled it to attract vast funds but the reality is that the US grid system, absence of pedestrians and cyclists and lack of jay walkers is not remotely representative of the world, it is in reality highly anomalous. Even on such young and simplified networks with very limited variance in road users they aren't getting autonomy to work with vehicles hitting lorries, static cars, walls and obviously that poor sod on a bike.
The real tests and development are the programs being runnin places like London and Oxford, old, complex systems with a huge array of different forms of transport all blended in mixed together. Even Oxford is far more formulaic and reliable than London as the locals whthwrnon foot, bike or car are more law obiding and less self centred.
It's the cracking of old metropolis such as London that will signify the viability of autonomy and I don't think that's happening anytime soon.
That echos my thoughts. Also, I don’t doubt there is a market for a true ADSA Level 5 autonomous car, where as you say people can go the pub and have it bring them home again. The problem with that, you picked up on, the car has to be so risk adverse that it gets to the point of not moving anywhere in an urban environment, or somebody has to be responsible in the event of a crash. You can bet that the manufacturers will ensure that the legislation is written so that is not them or their employees that are responsible, it will be the ‘driver’. Which means some level of manual control has to remain, which immediately removes ADSA Level 5 from the picture. The McKinsey graph earlier on makes no mention of level 5 cars and goes out to 2040.The fundamental flaw with autonomy is that as a pedestrian, once you know that the large, hurtling, leathal metal box is fully programmed to halt before your magnificence then those cars will never move in the blended urban environment. The only reason that pedestrians or cyclists give way to cars is because the car represents a big safety variable and therefor risk. Once the car is programmed to be reliably subservient then man is released from that fear to take advantage of this new freedom.
In shabbier parts of London you have always had the local tough guy who will select a vulnerable looking road user, make eye contact with the driver and step out into the road, forcing the driver to stop for them. Apply alcohol and the number of men who reveal themselves as that sort of person climbs steeply.
In places like Oxford St with high pedestrian footfall, when trying to cross from North to South or vice verse the traffic lights are largely irrelevant as the pedestrians keep crossing until the lead car specifically moves forward at them. It is only that movement that triggers an alert to pay attention to the rules. You already see that when the lead car is a tourist, not a resident, they end up never crossing until enough cars behind are blowing their horns. Cars programmed to not specifically drive at pedestrians will just sit there all day.
And the City? Home of the alpha male and wannabe alpha male. Once cars are programmed to give way then they will never move again.
As for any other part of London, I've already got used to Tesla's coming to a halt in the road because a pedestrian is waiting between two parked cars to cross and has, as always, positioned themselves forward of the line of parked cars to help vision and to shorten the crossing space.
Even on motorways which are arguably the simplest roads in the UK you can see the current tech isn't yet ready to cope with the huge array of curveballs and the dynamic environment that even the least powerful human brain can handle really quite easily.
https://www.reuters.com/article/us-uber-atg/uber-s...
The news that Uber is offloading it's autonomous division from its balance sheet just confirms how far away we are and remain. They've taken it off their books because it is not going to deliver revenue within any foreseeable economic timeframe. And in order to do that they've had to create a new entity to 'buy' it from them and to spread the ongoing capital drain.
Autonomy has had a hugely US centric mindset which has enabled it to attract vast funds but the reality is that the US grid system, absence of pedestrians and cyclists and lack of jay walkers is not remotely representative of the world, it is in reality highly anomalous. Even on such young and simplified networks with very limited variance in road users they aren't getting autonomy to work with vehicles hitting lorries, static cars, walls and obviously that poor sod on a bike.
The real tests and development are the programs being runnin places like London and Oxford, old, complex systems with a huge array of different forms of transport all blended in mixed together. Even Oxford is far more formulaic and reliable than London as the locals whthwrnon foot, bike or car are more law obiding and less self centred.
It's the cracking of old metropolis such as London that will signify the viability of autonomy and I don't think that's happening anytime soon.
I think the idea of people not owning a car, summoning a driverless pod to their location, and have them take it where they want to go, is at least 20+ years away, if ever.
The "responsible in the event of a crash" thing - do you mean criminally responsible, or do you mean financially liable? I am quite sure that even a level 5 car would require the keeper to be insured against third party risks caused by his use of the car on the road, so that wouldn't require any change. In the case of criminal responsibility, do you mean that it will be necessary for a passenger in a level 5 car to be able to order the car to drive dangerously?
otolith said:
The "responsible in the event of a crash" thing - do you mean criminally responsible, or do you mean financially liable? I am quite sure that even a level 5 car would require the keeper to be insured against third party risks caused by his use of the car on the road, so that wouldn't require any change. In the case of criminal responsibility, do you mean that it will be necessary for a passenger in a level 5 car to be able to order the car to drive dangerously?
I am thinking along the lines of in the 21st century, there is no such thing as an accident, RTA’s are now known as RTC’s for example.Whether somebody has to be responsible for criminal reasons or financial reasons, or just somebody to lay the blame on, means little in the world today. Somebody has to be responsible, and as soon as you get to that situation it doesn’t matter why.
Yes, financial liability can be address by insurance, but again somebody has to be a fault for the appropriate insurance to pay out. Meaning somebody carries the can, and it won’t be manufacturers, it will be drivers. But it can’t be the driver if there is no option for manual control, ergo ADSA Level 5 will not happen.
All of this my opinion obviously and could highly likely proven wrong, but the way the world addresses liability these days means somebody is always at fault, and it is generally the little person. That is not to say I agree with that, it’s just a sad reflection of the modern world, that seems to have started in America.
Edited by Megaflow on Tuesday 8th December 13:37
I read somewhere that a justification would be to save people money, as the asset on their driveway depreciating is only used x hours a day and therefore the rest of the time is inefficient.
I've got loads of stuff I don't use that much, but I just like owning it. It seems like the world is being drained of fun lately..
I've got loads of stuff I don't use that much, but I just like owning it. It seems like the world is being drained of fun lately..
jimPH said:
I read somewhere that a justification would be to save people money, as the asset on their driveway depreciating is only used x hours a day and therefore the rest of the time is inefficient.
I've got loads of stuff I don't use that much, but I just like owning it. It seems like the world is being drained of fun lately..
Cars tend to depreciate on both age and mileage so having it used more will still mean it will depreciate, but I agree we choose our cars to enjoy them not to save money. We could buy a cheap box and save a fortune or stick with what we have and enjoy it, don’t really want to share that with someone else.I've got loads of stuff I don't use that much, but I just like owning it. It seems like the world is being drained of fun lately..
DonkeyApple said:
The fundamental flaw with autonomy is that as a pedestrian, once you know that the large, hurtling, leathal metal box is fully programmed to halt before your magnificence then those cars will never move in the blended urban environment. The only reason that pedestrians or cyclists give way to cars is because the car represents a big safety variable and therefor risk. Once the car is programmed to be reliably subservient then man is released from that fear to take advantage of this new freedom.
If I was a teenager again, I would spend my evenings dashing out in front of autonomous cars and force an emergency brake to annoy the passengers (who have suffered several similar incidents already on the one trip). I wonder what they will call this new "teenage leisure" activity. Suggestions? There’s a fundamental and simple problem with the ownerless model. It’s called peak demand.
Today I buy a car and use it about 2% of the time, so surely it is possible to use the car more intensively? One car between 10 people could make someone rich. The issue is that when I buy a car, I’m not just paying for transport, I’m paying for convenience. Convenience is:
- the car is outside my house now. Not in 10 minutes.
- I can change my mind at the last minute - nope, I’m not going into town right now
- I can leave stuff in the car, and expect it to still be there the next day
Peak demand is why you can’t get a cab on Saturday night, and why you can’t get a cab on Monday morning to go to work. The delta between off peak and peak limits the size of the market. You’re not going to have hordes of cab drivers springing into action on Saturday night if they can’t get work during the week.
In the same way, any shared ownership model has to be limited by the delta between peak and off peak. The idea that someone can make hundreds of k out of a car is a calculation by a spotty consultant who has taken the uber hourly rate and multiplied it by the number of hours in a week.
The other issue is that the current solution is so damn cheap. Aside tax and insurance, I can get a slice of that convenience for £500. If I want fancier convenience, I can pay £500 a month. Between these extremes there are thousands of options. Any shared ownership model has to be better than this - either more convenient, or cheaper. That’s a tough sell.
Today I buy a car and use it about 2% of the time, so surely it is possible to use the car more intensively? One car between 10 people could make someone rich. The issue is that when I buy a car, I’m not just paying for transport, I’m paying for convenience. Convenience is:
- the car is outside my house now. Not in 10 minutes.
- I can change my mind at the last minute - nope, I’m not going into town right now
- I can leave stuff in the car, and expect it to still be there the next day
Peak demand is why you can’t get a cab on Saturday night, and why you can’t get a cab on Monday morning to go to work. The delta between off peak and peak limits the size of the market. You’re not going to have hordes of cab drivers springing into action on Saturday night if they can’t get work during the week.
In the same way, any shared ownership model has to be limited by the delta between peak and off peak. The idea that someone can make hundreds of k out of a car is a calculation by a spotty consultant who has taken the uber hourly rate and multiplied it by the number of hours in a week.
The other issue is that the current solution is so damn cheap. Aside tax and insurance, I can get a slice of that convenience for £500. If I want fancier convenience, I can pay £500 a month. Between these extremes there are thousands of options. Any shared ownership model has to be better than this - either more convenient, or cheaper. That’s a tough sell.
snowandrocks said:
Cleanliness is one issue with the ownerless model that is rarely mentioned.
I worked as a private hire driver whilst at uni and if there's one lesson I learnt, it's that human beings are disgusting! Add alcohol to the mix and they become really disgusting!
I had quite a nice A6 to drive around and did my best to keep it clean but it was a never ending thankless task. Between almost every job I was disposing of rubbish, every 3 or 4 jobs I'd be picking up dropped chips and wiping up sauce from spilled kebabs. That's all before we get to the bodily fluids. The car then needed properly cleaned, hoovered and wiped down after every shift.
Now, if they behave like that with the owner of the car sitting next to them asking them not to eat, how are they going to behave if they're on their own?
Something I mentioned at the time in the thread. However, Uber would probably base a customer rating on how many turds, bogies, vomits and semen Staines they left in one of their utility pods. I worked as a private hire driver whilst at uni and if there's one lesson I learnt, it's that human beings are disgusting! Add alcohol to the mix and they become really disgusting!
I had quite a nice A6 to drive around and did my best to keep it clean but it was a never ending thankless task. Between almost every job I was disposing of rubbish, every 3 or 4 jobs I'd be picking up dropped chips and wiping up sauce from spilled kebabs. That's all before we get to the bodily fluids. The car then needed properly cleaned, hoovered and wiped down after every shift.
Now, if they behave like that with the owner of the car sitting next to them asking them not to eat, how are they going to behave if they're on their own?
jimPH said:
I read somewhere that a justification would be to save people money, as the asset on their driveway depreciating is only used x hours a day and therefore the rest of the time is inefficient.
I've got loads of stuff I don't use that much, but I just like owning it. It seems like the world is being drained of fun lately..
Back in the day of traditional minicabs there were plenty of people in London who had the family S class pay its way by having the cousin who lived in the garden she'd drive it through the night. What the market is really suggesting is that as people shop themselves into poverty it won't just be the new arrivals who have to sweat their assets to make ends meet. I've got loads of stuff I don't use that much, but I just like owning it. It seems like the world is being drained of fun lately..
In short, no one really wants to rent out their mattress by the hour while they are at work but more and more people need to do so if they want to have their consumer luxuries
Edited by DonkeyApple on Tuesday 8th December 22:46
Automated cabs will have to have variable pricing like Ubers surge to profit people from taking them when it's raining but be cheap enough when it's dry and sunny that people hop in for a minute rather than walk five.
DonkeyApple said:
I
The fundamental flaw with autonomy is that as a pedestrian, once you know that the large, hurtling, leathal metal box is fully programmed to halt before your magnificence then those cars will never move in the blended urban environment. The only reason that pedestrians or cyclists give way to cars is because the car represents a big safety variable and therefor risk. Once the car is programmed to be reliably subservient then man is released from that fear to take advantage of this new freedom.
It won't just be pedestrians it will be other car drivers and cyclists as well knowing that the driverless car will yield no matter what.The fundamental flaw with autonomy is that as a pedestrian, once you know that the large, hurtling, leathal metal box is fully programmed to halt before your magnificence then those cars will never move in the blended urban environment. The only reason that pedestrians or cyclists give way to cars is because the car represents a big safety variable and therefor risk. Once the car is programmed to be reliably subservient then man is released from that fear to take advantage of this new freedom.
DonkeyApple said:
jimPH said:
I read somewhere that a justification would be to save people money, as the asset on their driveway depreciating is only used x hours a day and therefore the rest of the time is inefficient.
I've got loads of stuff I don't use that much, but I just like owning it. It seems like the world is being drained of fun lately..
Back in the day of traditional minicabs there were plenty of people in London who had the family S class pay its way by having the cousin who lived in the garden she'd drive it through the night. What the market is really suggesting is that as people shop themselves into poverty it won't just be the new arrivals who have to sweat their assets to make ends meet. I've got loads of stuff I don't use that much, but I just like owning it. It seems like the world is being drained of fun lately..
In short, no one really wants to rent out their mattress by the hour while they are at work but more and more people need to do so if they want to have their consumer luxuries
Edited by DonkeyApple on Tuesday 8th December 22:46
Gassing Station | General Gassing | Top of Page | What's New | My Stuff