We'd like to remind Forumites to please avoid political debate on the Forum... Read More »
Debate House Prices
In order to help keep the Forum a useful, safe and friendly place for our users, discussions around non MoneySaving matters are no longer permitted. This includes wider debates about general house prices, the economy and politics. As a result, we have taken the decision to keep this board permanently closed, but it remains viewable for users who may find some useful information in it. Thank you for your understanding.
📨 Have you signed up to the Forum's new Email Digest yet? Get a selection of trending threads sent straight to your inbox daily, weekly or monthly!
Fully automated vehicles - 'not in our lifetime'?
Options
Comments
-
So for instance if I look out onto my street right now there are maybe 60 cars parked up in drives and on the street.0
-
For a taxi to turn up in 5 seconds you'd need one every 25m and hope no-one else needs one - total gridlock.
You sound like a horse and carriage driver in the 1800s saying that humans aren't built to travel at 30mph.
Self-driving cars could be 25cm apart and they will never be gridlocked. The one thing a self-driving car already does much better than humans is to not gawp at a line of moving traffic ahead of them while sitting completely still, until the gawper in front of them finally moves. If the entire road is filled with self-driving cars, all of them will move simultaneously, and gridlock can never occur.
When a traffic light goes green in front of a queue of human motorists, it can take minutes between the lights going green and the human at the back to start moving because of the need for each motorist to scratch their balls while they wait for the car in front to move. In a queue of self-driving vehicles, they will all set off simultaneously. Of course, there won't be any traffic lights as all cars will merge seamlessly at junctions.
Self-driving cars can never gridlock, a traffic jam in self-driving cars terms means cars dropping their speed slightly for a few seconds to create space for an upcoming merge.
This will require human drivers to be banned from the roads but this is GreatApe's utopian vision of 2050 we're talking about (I think GreatApe uses the Islamic calendar).0 -
In a queue of self-driving vehicles, they will all set off simultaneously.0
-
Just because we have 32 million cars now doesn't mean 32 million robo cars will get one to anyone in seconds. It's unnecessary anyway as ordering it and walking outside will take longer that that. Like I said, 5 minutes is fine if you can rely on it.
You seem to be talking of everyone having their own private car but leasing it instead of using a robo taxi. So maybe we're talking at cross purposes. Whoever owns the cars isn't going to want them sitting idle unless they are charging and having them sitting waiting for fares unless they know they are about to be summoned. Especially if there are different options regarding seats and luggage.
The new system just needs to be better than the current system
The current system is cars do about 10,000 miles a year the new system will be that robot taxis do 30,000-50,000 miles a year but we have 1/3rd to 1/5th the current stock of cars.
You won't even have to hail them you just walk outside and get into one that's parked there. Like I said on a street like mine there will always be a dozen or more just parked waiting.
They will still be cheaper and more profitable than the current system.0 -
But with less cars it's less likely there will be one right there.
I largely agree with you, but I don't think they need to be that responsive.
On the gridlock - sure they'll be able to navigate round themselves with maybe 10-50cm to spare, but it's still basic physics in that the more of them on the road the more they'll need to compensate. 10,000,000 of them will be able to travel around much faster than 30,000,000, especially since they'll still need to stop to let humans board or cross roads.
If cheap enough, they'll take over a lot of journeys, but they won't be replacing the private car until the private car becomes too expensive for the majority to own.0 -
The main problems are going to be trying to integrate auto & non-auto vehicles and segregation between vehicles & "non-vehicles" It may be "easy" to predict what a fully auto car is going to do but very near impossible with a human driver. Also, until AI knows what to do with pedestrians & objects in the road, it's a non-starter. What will it do with pedestrians walking out in front of it? Will people wait to cross the road or just step out knowing that the vehicle will stop for them? What about children running into the road? How will it deal with cats/dogs/pigeons/rabbits etc (pigeons especially where I live as they wait until the very last moment before flying out of the way). If it's seriously on the horizon, why haven't they done the easy stuff first? Trains run on tracks & cannot deviate from their course. They are kept separate from pedestrians & stop/start in specific places. Why aren't they all driverless? Why aren't planes pilotless? Most of the same segregation arguments apply. They are separated from each other by miles (and have an extra dimension to work in) So, again, why hasn't it been done on a major scale? And ships sailing in the vastness of the oceans? But no, lets concentrates on cars that are inches apart, run on roads with junctions every few yards and road surfaces who's conditions change depending on the weather ect. & are closely mixed up with soft squishy people? But I'm probably just a Luddite!0
-
But with less cars it's less likely there will be one right there
It will be such that there is one right there. You can shrink the stock drastically and that would still be the case. Right now outside on this road there are at least 30 cars parked within a 30 second walk. In the future that might be 10 cars outside just waiting.I largely agree with you, but I don't think they need to be that responsive.
To begin with they will operate like uber taxis but as the numbers grow they will look and act more like private cars in that every street will have some parked up waiting so you never really have to wait just walk up to it and get in.On the gridlock - sure they'll be able to navigate round themselves with maybe 10-50cm to spare, but it's still basic physics in that the more of them on the road the more they'll need to compensate. 10,000,000 of them will be able to travel around much faster than 30,000,000, especially since they'll still need to stop to let humans board or cross roads.
I dont think they will drive close to each other that would be too risky a mechanical failure instead of causing one crash might cause a 100 fleet tailgating crash.
It will either be shared fleets to reduce the road congestion or the fleets will mostly be 1 seater cars half the size of a smart car with some 2 seaters and much less 4/6/10 seaters.If cheap enough, they'll take over a lot of journeys, but they won't be replacing the private car until the private car becomes too expensive for the majority to own.
They wont just be cheaper they will be more convinient
We got rid of the second car when uber arrived, we would probably get rid of the main car if robo taxis come to being. Its not just about the cost it also saves time and hassle. Who likes taking time out to compare and buy insurance, to go view half a dozen cars to buy one, to put up your old car for sale, to deal with breakdowns maintenance road tax etc and of course parking problems too.
Would be much easier just to use the robo taxis or for when you need it rent them for a day or week or month as a temporary private use car0 -
Truffle_Snuffler wrote: »The main problems are going to be trying to integrate auto & non-auto vehicles and segregation between vehicles & "non-vehicles" It may be "easy" to predict what a fully auto car is going to do but very near impossible with a human driver. Also, until AI knows what to do with pedestrians & objects in the road, it's a non-starter. What will it do with pedestrians walking out in front of it? Will people wait to cross the road or just step out knowing that the vehicle will stop for them? What about children running into the road? How will it deal with cats/dogs/pigeons/rabbits etc (pigeons especially where I live as they wait until the very last moment before flying out of the way). If it's seriously on the horizon, why haven't they done the easy stuff first? Trains run on tracks & cannot deviate from their course. They are kept separate from pedestrians & stop/start in specific places. Why aren't they all driverless? Why aren't planes pilotless? Most of the same segregation arguments apply. They are separated from each other by miles (and have an extra dimension to work in) So, again, why hasn't it been done on a major scale? And ships sailing in the vastness of the oceans? But no, lets concentrates on cars that are inches apart, run on roads with junctions every few yards and road surfaces who's conditions change depending on the weather ect. & are closely mixed up with soft squishy people? But I'm probably just a Luddite!
Trains and planes dont need to be pilot less because the cost of the pilot per passenger mile is low much lower than for a taxi where the taxi human driver is perhaps as much as 70% of the cost of a taxi trip
As for ships to rent a cape sized vessel can cost more than a hundred thousand dollars per day while the captain is paid maybe $300 a day and thus is again a trivial cost. If human wages went up towards $50,000 a day then we would probably automate ships trains and planes too
While the challenge is very difficult the rewards are massive.
The company that brings out self drive car software will be worth >$10 trillion it also has massive economic productivity improvements so much so that the whole worlds fortunes will change.
Before AI I would have also said it was impossible for humans to program every single possible scenario but with AI it can drive a billion human years and program itself by seeing any and all imaginable scenarios. Also it does not need to be perfect even if it kills a million people each year that is still an improvement over human drivers.0 -
Another problem is trying to use a new system on existing infrastructure. No effective segregation of traffic & non-traffic and roads that have 90 degree turns to get in & out of side streets which slows or stops the traffic behind every few yards. The real way to make it feasible would be for the new system to be completely separate from the old.0
-
This discussion has been closed.
Confirm your email address to Create Threads and Reply

Categories
- All Categories
- 351.2K Banking & Borrowing
- 253.2K Reduce Debt & Boost Income
- 453.7K Spending & Discounts
- 244.2K Work, Benefits & Business
- 599.3K Mortgages, Homes & Bills
- 177K Life & Family
- 257.6K Travel & Transport
- 1.5M Hobbies & Leisure
- 16.2K Discuss & Feedback
- 37.6K Read-Only Boards