Could driverless cars replace railways?

[90019]

The list of point against driver less cars is almost endless.

This.

We've been over the whole driverless cars thing many times before, and every time the idea gets completely shot down.
I'd go over it again, but I can't be bothered just now.

One thing I do wonder is how those for whom life behind the wheel is as much a reason for being as life on the iron road is for us, will react to being cocooned in a computer driven car !

Badly.

If the roads in this country were to change to automated vehicles, I'd emigrate.



Another point - if we automate all the vehicles on the roads, what do you propose doing with all the people like myself who drive professionally for a living?

 

[yorksrob]

imagine the reception the idea of Ryanair introducing pilot less aircraft would get.

Well there is that. I'm not a fan of flying, and even on a modern "fly by wire" I'd rather there were still a pilot just in case !
--- old post above --- --- new post below ---

Badly.

If the roads in this country were to change to automated vehicles, I'd emigrate.

Pretty much as I suspected then

 

[Muzer]

I dont think asking how the system would cope with multiple high speed motorway pile ups is a trivial reason.

That's not what I was referring to

Another point - if we automate all the vehicles on the roads, what do you propose doing with all the people like myself who drive professionally for a living?

Same thing as guards on the London Underground, I expect would be the answer of some

 

[deltic]

I have to agree with the OP. Except for the largest cities where congestion will still limit use of cars and perhaps for long distance journies public transport as we know it is likely to disappear within 50 years. Remaining trains and buses will of course also be driverless. Given road vehicles are responsible for some 3,000 deaths a year it is highly likely that people will be banned from actually driving once the technology is fully proven.

There is of course no need to own a car you will just be able to book one as you would a minicab but the cost would be about 50% cheaper as you dont have to pay for a driver and can book a car with just as many seats as passengers.

The big if of course is that this will only work if the cost of the technology required can be got down to roughly the present cost of a car.

 

[samxool]

I think the biggest thing holding back driverless cars besides technology and infrastructure, is insurance. Are the big car manufacturers going to take over the.insuring of driverless cars, wait for the inevitable system failure that leads to a massive pile up?
nope, humans controlling cars, thus taking on the burden of control, thus the burden of car insurance is here to stay!

 

[radamfi]

Driverless cars could make it easier for people to use trains. There would be no need for local buses or expensive manned taxis to get to the station. Driverless taxis or small minibuses could pick up passengers to and from the station based on real time demand. You could have the choice of a taxi to yourself or a shared taxi at lower cost.

 

[HSTEd]

Driving around in Arizona, SoCal or Nevada is one thing.

How does the system cope with near zero visibility?
Hundreds of non cooperative radars pinging away at random is never going to be a good idea which means you have serious sensor issues.

And then we have to include the fact that many people would not like to be cooped up in a tiny vehicle for hours on end unable to move significantly while crawling along at only 60-70mph.

And what happens if someone starts jamming radars used for poor weather sensory inputs? You could get massive pileups.
It is very hard to remotely, electronically, screw up a railway system such that lots of people die.

 

[yorksrob]

I have to agree with the OP. Except for the largest cities where congestion will still limit use of cars and perhaps for long distance journies public transport as we know it is likely to disappear within 50 years. Remaining trains and buses will of course also be driverless. Given road vehicles are responsible for some 3,000 deaths a year it is highly likely that people will be banned from actually driving once the technology is fully proven.

There is of course no need to own a car you will just be able to book one as you would a minicab but the cost would be about 50% cheaper as you dont have to pay for a driver and can book a car with just as many seats as passengers.

The big if of course is that this will only work if the cost of the technology required can be got down to roughly the present cost of a car.

Well, your argument does pre-suppose that only Cities and motorways suffer from congestion. As we know, even with the level of car ownership that we have at present, there is still congestion in many other areas of the country, so unless we suffer a catastrophic collapse in population (which again isn't inconceivable, but we'll leave that unhappy possibility to one side) there would still come a crunch where we would have to choose between more public transport or more tarmac.

Then energy. Unless we crack the energy issue, fuel will still need to be paid for and might end up comparatively even more expensive than now with a growth in world population, so we could still end up with a greater need for public transport.

 

[deltic]

This.


Another point - if we automate all the vehicles on the roads, what do you propose doing with all the people like myself who drive professionally for a living?

probably 90% of the jobs we do now didnt exist 100 years ago and 90% of them wont exist in another 100 years ago -

 

[carriageline]

I think nym needs to have a read/discussion on this, he is clued up with sensors etc, and explained why we can't have a fully automatic tube, running on sensors.

How would sensor fatigue affect these? And as far as I understand it, a computer cannot react faster than us, yet either.

IIRC, a massive technology that provide us with a great way to run cars (might be electric, or some alternative, I cannot remember), one of the oil giants bought the rights to it, so we will probably never see it. I can't remember a great deal though I'm afraid..

 

[Sun Chariot]

30 years ago, I was using a BBC model B computer to...join up points and make a circle. So it's highly likely we'll have a more joined-up infrastructure for our autonomous cars, vans, buses, trucks, motorbikes and mobility-scooters.

I don't agree, though, that the costs and constraints of personal transport ownership will remove need for our passenger and freight public transport services.

I also suspect that, over the next 50 years, private (vehicle) travel costs will become more prohibitive to even more people, compared to public travel cost over an equivalent route.

 

[MarkyT]

Just think of the potential benefits of driverless cars:

+ No need for a driving licence.

Doubtful - only possible if there's no manual override and even then there may be some basic competence skills required in order to get insurance.

. . .

+ Going to town and telling to car to drive home to save on parking costs.

Really? twice the total distance travelled would be very inefficient. Cars could drive autonomously to remote car parks however, still fairly close by but in less dense areas

+ Cars to driving closer together at higher speeds, increasing road capacity.

Or alternatively further apart closer to actual safe stopping distance rather than high speed tailgating.

+ No more drunk driving and fewer road traffic accidents, so lower insurance costs.

Only once there are no more humans driving on the road at all

+ Battery-powered, so very low fuel costs.

Not specific to autonomy

+ Autonomous car hire pools revolutionising urban transport.

Robocabs could be a very disruptive technology for urban areas with major implications for the future of conventional taxis and busses. They could work very well together with fast interurban and intercity train services

. . . Railways were born into a world of streetcars, trolleybuses, narrowboats, horse-draw carriages and telegrams. They are the last man standing from that Victorian age - and everything has its time. Are railways doomed?

The modern motor vehicle is little more than a high tech farm cart, an invention born into a world of ox and horse haulage many hundreds even thousands of years before widespread adoption of the rail road. Easier and cheaper automation technology could also reap benefits in the rail industry, changing the optimum vehicle size and extending operating hours and increasing service frequency whilst reducing costs. Transport technologies rarely 'have their day' and expire, rather they adapt and their relative niches adjust.

 

[deltic]

Well, your argument does pre-suppose that only Cities and motorways suffer from congestion. As we know, even with the level of car ownership that we have at present, there is still congestion in many other areas of the country, so unless we suffer a catastrophic collapse in population (which again isn't inconceivable, but we'll leave that unhappy possibility to one side) there would still come a crunch where we would have to choose between more public transport or more tarmac.

Then energy. Unless we crack the energy issue, fuel will still need to be paid for and might end up comparatively even more expensive than now with a growth in world population, so we could still end up with a greater need for public transport.

driverless cars will easily increase usable capacity on the road network twofold significantly reducing congestion in most towns.

running costs will depend greatly on the model adopted for ownership of driverless cars - one model suggests people will own their cars as now and hence purchase bigger vehicles than needed most of the time - the other model is that we will use them more as taxis and hire the appropriate sized car - single seater to go to work - 12 seater for a night out with the mates - camper van for an overnight trip across the country. Which would reduce costs.

I agree energy costs may be an issue - my own view is that as we move away from petrol/diesel and hence fuel duties, road user charging will have to come in that will limit demand for car traffic generally.

 

[jon0844]

I think there will be a lot of resistance to driverless cars from people that like to drive.

Why? Ultimately for such systems to work well and at higher speeds, there will need to be legislation to stop manually driven cars altogether (well, perhaps except emergency vehicles that will communicate their location to surrounding cars and stop them).

It just isn't going to work on open roads with varying conditions, pedestrians, animals, obstructions (that you or I might steer around) and many more things we could all think of in the next five minutes.

It isn't the same as a system to control a train or plane in tightly controlled conditions.

If anything, getting rid of manually driven cars in all but off road/track locations might get more people using public transport.

Technically such systems are feasible, perhaps inevitable, but I really don't think the two will coexist and if they do, driverless cars will be far too restricted for safety reasons.

 

[yorksrob]

driverless cars will easily increase usable capacity on the road network twofold significantly reducing congestion in most towns.

running costs will depend greatly on the model adopted for ownership of driverless cars - one model suggests people will own their cars as now and hence purchase bigger vehicles than needed most of the time - the other model is that we will use them more as taxis and hire the appropriate sized car - single seater to go to work - 12 seater for a night out with the mates - camper van for an overnight trip across the country. Which would reduce costs.

I agree energy costs may be an issue - my own view is that as we move away from petrol/diesel and hence fuel duties, road user charging will have to come in that will limit demand for car traffic generally.

A more efficient use of road capacity would undoubtedly have an effect on reducing road congestion. However, I'm not convinced that it would be by such a great extent.

When the passenger gets to their destination, is it stored, in which case, you still need parking, or does the car return back home, in which case it makes twice the journeys, cancelling out congestion benefits and increasing running costs.

The concept of people hiring cars as taxis is an interesting one. However, I think one of the main benefits of car travel for many people is the sense of a personal domain, so my own hunch is that people won't go for that.

I think that a breakthrough on the energy front car wise, combined with widespread adoption of driverless technology would have me hitting the gin though.

 

[jon0844]

How are the various car hire schemes doing in London where you can jump into a car parked nearby?

Bikes are doing fine, clearly, but I'm not sure about the car schemes.

 

[90019]

the other model is that we will use them more as taxis and hire the appropriate sized car - single seater to go to work - 12 seater for a night out with the mates - camper van for an overnight trip across the country. Which would reduce costs.

No it wouldn't.
Consider it this way - how many of every different type of vehicle would you have to have to be able to meet the possible demand at any one time?

How long would you have to wait for your autonomous car to arrive?
Where would you store them all and how many of each type would you have to keep in each place?
How much wasted milage are you going to end up having with all these autonomous cars coming from wherever they're stored to where they're needed?


Right now, both my cars are at my door and depending on what I want to do, I can walk outside, get into whichever one is more appropriate (or whichever one I feel like driving) and set off immediately.
If I have to wait on an autonomous car arriving, how long do I have to wait for the right one to be sorted out and to arrive here to then take me where I want to go?

 

[deltic]

No it wouldn't.
Consider it this way - how many of every different type of vehicle would you have to have to be able to meet the possible demand at any one time?

How long would you have to wait for your autonomous car to arrive?
Where would you store them all and how many of each type would you have to keep in each place?
How much wasted milage are you going to end up having with all these autonomous cars coming from wherever they're stored to where they're needed?


Right now, both my cars are at my door and depending on what I want to do, I can walk outside, get into whichever one is more appropriate (or whichever one I feel like driving) and set off immediately.
If I have to wait on an autonomous car arriving, how long do I have to wait for the right one to be sorted out and to arrive here to then take me where I want to go?

As I said there are two basic models - one you own a car so it is no different to your present position. The other is that you hire one as needed. You only need 4-5 basic car models - plus some specialised vehicles - no more than hire car companies have now. In most urban areas you could easily have cars at most only 3-4 miles away or less than 10 minutes - easily bookable online. I agree there would be more empty mileage than now but for those doing a low mileage it would be far cheaper than owning a car
--- old post above --- --- new post below ---

How are the various car hire schemes doing in London where you can jump into a car parked nearby?

Bikes are doing fine, clearly, but I'm not sure about the car schemes.

Not that well - once people sign up to the scheme they rapidly reduce their car use as they realise how much each trip actually costs.

Bike hire doesn't work particularly well either because of the cost of moving bikes around

 

[Eagle]

You only need 4-5 basic car models - plus some specialised vehicles - no more than hire car companies have now.

No more models maybe, but you'd need a lot more total vehicles than a current car hire company. Hundreds of times more in fact.

 

[Sun Chariot]

How are the various car hire schemes doing in London where you can jump into a car parked nearby?

Bikes are doing fine, clearly, but I'm not sure about the car schemes.

Last time I read one of the ads on the tube (a year ago) they charged between £4 and £5 per hour. Fine for the quick trip but that's more than £100 for a 24 hour hire period.

 

[deltic]

No more models maybe, but you'd need a lot more total vehicles than a current car hire company. Hundreds of times more in fact.

there are over 30m cars in the UK - not sure what the maximum number on the roads is at any one time - possibly 20-25m. Some people therefore argue that if everyone hired a vehicle fewer would be needed.

Not sure the car ownership model is that relevant to whether and how driverless cars will impact on public transport network or not.

Most new cars are paid for by finance companies so doesnt really matter if they are brought by individuals or car hire companies.

 

[radamfi]

If driverless cars are proven to be safer than manually driven ones, it would be hard to justify the continued use of normal cars if it is seen that people are needlessly being injured or killed. I suppose there could be private racetracks built around the country for enthusiasts.

 

[Eagle]

there are over 30m cars in the UK - not sure what the maximum number on the roads is at any one time - possibly 20-25m. Some people therefore argue that if everyone hired a vehicle fewer would be needed.

I think it'd possibly be the other way round under your scheme, if you plan on having a lot of 1- and 2-person cars; that would act to stop people sharing cars and more people to make journeys on their own.

 

[90019]

If driverless cars are proven to be safer than manually driven ones, it would be hard to justify the continued use of normal cars if it is seen that people are needlessly being injured or killed.

It'd be a real shame if that were to happen, as there are some fantastic driving roads around the UK.

 

[yorkie]

Railways were born into a world of streetcars, trolleybuses, narrowboats, horse-draw carriages and telegrams. They are the last man standing from that Victorian age - and everything has its time.

Railways are infrastructure, the other items you list are forms of transport, that require infrastructure.

The fact that horses were used to haul road and rail vehicles when railways were created, is just as irrelevant as saying that horses were used when roads were created.

There are also fundamental flaws in this proposal, for example when a car turns out, it is meant to use a sensor to detect a gap in traffic? Yet traffic will be much closer together, apparently... so it could take a lot longer to turn from a side road into a main road.

The idea is flawed, unaffordable and unsustainable in my opinion.

 

[David Barrett]

They were doomed at the hands of the private car, going to be ousted by cheap efficient helicopter services, completely eradicated by fixed wing airline competition, put out of business by cheap coach travel and trodden into the ground by highly competitive deregulated stage carriage bus operations, in much the same way as they will become history because of this new technology whose shortcomings have not, as yet, been forseen by those who wish to promote the new promised era of cheap bunk.

 

[jon0844]

If you made every car computer controlled, using GPS, GLONASS or whatever (and God help us if any of these systems ever went down as it would effectively grind the entire country to a halt!) then a car could merge with traffic, or change lanes easily, as other cars would be made to stop. Cars would communicate with the others.

Ordinary motorists wouldn't automatically let a computer controlled vehicle in, so driverless cars would no doubt have to emergency stop in such situations, as well as going slow in the first place (to ensure the passengers aren't thrown around too much) and then working out how to merge with fast moving traffic - no doubt taking far longer than ordinary drivers.

Computer control COULD let cars go a lot faster and closer together, but there will probably be many regulations to stop this from actually happening. I actually expect we'd see cars driving slower, until such time that you built special roads free of uncontrollable risks, like pedestrians and animals. You'd probably be fencing these off, like motorways or indeed railway lines.

If roads were also kept clear of parked vehicles and other obstructions, this also begs the next question; where would all these vehicles stop/park when not in use?

It would mean a massive spend in infrastructure to allow for all of this to happen, while many motorists would be very resistant. Otherwise, driverless cars will NOT be simply allowed to drive as normal amongst everything/everyone else.

I expect that we'll see cars controlled by a driver, as now, but with a stack of technology to make it easier to drive - including cars able to communicate with each other and avoid collisions (but only ever stepping in to stop a collision, not actually drive the car). A lot of that technology exists right now, and can be bought on, say, the new Volvo V40 - but that can only stop itself, not another car (for now).

 

[rebmcr]

When a car turns out, it is meant to use a sensor to detect a gap in traffic? Yet traffic will be much closer together, apparently... so it could take a lot longer to turn from a side road into a main road.

The cars all talk to each other and share information.

They'll be able to reliably and safely pull into barely-existent gaps, and accelerate to the precise speed of the surrounding cars, in a manoeuvre that would cause the average person to evacuate their bowels.

They'll be able to travel at 120mph on motorways, with a half-second gap between vehicles, and (if necessary) all come to a rapid stop as one entity, collision-free.

They'll be able to weave around each other — at speed — when crossing paths on flat junctions. Every traffic light is green.

They'll be able to regain control and safely stop after a tyre blowout or other malfunction, in a fraction of the time a human being would.

They WON'T be able to spot that cyclist, or that child, or that ice-cream van and take appropriate action in the same way that a human being would. Not yet anyway — that sort of processing is starting to happen in labs, but has at least 15 years before it starts to sniff the air of commercial products.

 

[cmjcf]

There are also fundamental flaws in this proposal, for example when a car turns out, it is meant to use a sensor to detect a gap in traffic? Yet traffic will be much closer together, apparently... so it could take a lot longer to turn from a side road into a main road.

The idea is that the autonomous vehicles talk to each other to coordinate the flow. A vehicle approaching a junction might find itself effectively "flashed on" by the tenth vehicle on the main road some time before either have reached the junction.

 

[Cherry_Picker]

While I don't doubt that autonomous cars could make a real dent in certain parts of public transport, especially in rural and suburban areas (and I think they are a big threat to bus and coach transport too) they will struggle to cope with the sheer number of people a train has the capacity to handle. Waterloo handles over 100 million journeys a year, Victoria 85 million, Liverpool Street and London Bridge are both over 60 million. They say 300,000 people pour out of Waterloo station in the morning peak and the road network simply cannot handle that. Okay they might not converge on Waterloo station any more, but the City of London is one square mile and a similar number (300k) non residents work there too.

As a train driver it's the inevitable autonomous intercity/interurban express train I am more concerned about. It will happen, but I think (or rather, I hope) it won't happen until I'm retired!