Driverless cars could revolutionise transport and lead to the decline of Railways

[Trainer2]

With due respect, I don’t think you have considered the full potential of driverless cars.

125mph is very much achievable

My comments are from someone who is a rail enthusiast, you can’t beat a more convenient, safer, cheaper form of technology.

 

[lkpridgeon]

With due respect, I don’t think you have considered the full potential of driverless cars.

125mph is very much achievable

My comments are from someone who is a rail enthusiast, you can’t beat a more convenient, safer, cheaper form of technology.

Whilst 125mph is technically achievable, the safety would be a completely different issue to counter. It would have to be done on purpose built roads that are well maintained, reasonably straight, extended stopping distances, long distances between junctions, and relatively flat, sounds familiar eh? At which point you might as-well just build a modern high capacity railway such as HS2 and other smaller schemes.

The technology simply isn't currently here to make a viable driver-less car as mapping, location services, computer vision, computational resource (that's small and energy efficient enough) simply isn't here yet and won't be for a long while. Also regulatory, insurance, tax issues and other legal malarkey need sorting too of which will take a good while to iron out if the impetus is even there to do so.

At the moment from what I can see driverless cars are stuck in the needlessly expensive vanity project stage of development and we simply can't base future transport plans on a breakthrough with them happening any time soon. And when they do come round I see them acting more like taxi's that shuffle people to/from a mainline station should they be going a long distance where you pay per use instead of owning one.

 

[Annetts key]

With due respect, I don’t think you have considered the full potential of driverless cars.

Regardless of the benefits of fully autonomous cars, they are unlikely to have any significant positive impact on the traffic jams that many of our urban areas suffer from.

And even if they use electric motors rather than ICE, rubber tyres still cause some pollution.

 

[Trainer2]

Regardless of the benefits of fully autonomous cars, they are unlikely to have any significant positive impact on the traffic jams that many of our urban areas suffer from.

And even if they use electric motors rather than ICE, rubber tyres still cause some pollution.

Traffic Jams will not be an issue, vehicles controlled by a central computer will move as fast as physically possible.
This may result in walking pace in major cities but it will still be a huge improvement over existing car driving.
Plus there is no need for parking and walking to your final destination, the car will drive away and find it’s own parking space or next user.
this is far more convenient than a train.
Door to door, faster, cheaper and the luxury of your own private transport, no heading to a station, no waiting for a train that may be late, no sharing a carriage, no additional transport from your destination station to your end of journey location.

One needs to consider an autonomous car will look very different from a current car. It’s likely they will have a smaller foot print as no requirement for a lot of the driving controls.

long term the railway is finished with few exceptions.

 

[Annetts key]

Traffic Jams will not be an issue, vehicles controlled by a central computer will move as fast as physically possible.
This may result in walking pace in major cities but it will still be a huge improvement over existing car driving.

I still don’t see how you can get more road vehicles on the same road network all at the same time, as commuters travel to, or from work. Or people going on holiday or for the weekend fill the motorways during the summer holiday season.

And I call traffic moving at 10 MPH or less, a traffic jam.
Plus, not all vehicles will be autonomous, there will still be some vehicles driven by humans, because some people like driving.
And I very much doubt that there will be one central control computer.

 

[Trainer2]

You make a good point, it needs to be fully automated on the roads for it to work effectively

 

[vinnym70]

You make a good point, it needs to be fully automated on the roads for it to work effectively

And while there are any 'manually' driven cars on the roads, any hope of a sensible, safe, autonomous vehicle remains a pipe-dream. AI or not, computers find it very difficult to deal with anomalies and pretty much everyone who drives a car is an anomaly in some respect or other.

At some point in the future I imagine our 'roads' will behave more like an Ocado warehouse (it's very impressive and worth a watch on YouTube if you've not seen it) but while there's a single car under human control, what robotics can achieve isn't, in my opinion, attainable.

 

[SuperNova]

And while there are any 'manually' driven cars on the roads, any hope of a sensible, safe, autonomous vehicle remains a pipe-dream. AI or not, computers find it very difficult to deal with anomalies and pretty much everyone who drives a car is an anomaly in some respect or other.

At some point in the future I imagine our 'roads' will behave more like an Ocado warehouse (it's very impressive and worth a watch on YouTube if you've not seen it) but while there's a single car under human control, what robotics can achieve isn't, in my opinion, attainable.

And guess what, some of us humans enjoy driving... driverless cars are a nothing but a gadgetbahn. Railways will continue to provide an effective method of transportation for many decades, if not centuries to come.

 

[ivorytoast28]

And guess what, some of us humans enjoy driving...

The same way some humans enjoy flying... or sailing.. but in the end it always turns out that the majority would rather take the easy route. Sure many car lovers would love to race cars on cool tracks, but the average driver would give up their daily M25 "experience" if they could in an instance

 

[SuperNova]

The same way some humans enjoy flying... or sailing.. but in the end it always turns out that the majority would rather take the easy route. Sure many car lovers would love to race cars on cool tracks, but the average driver would give up their daily M25 "experience" if they could in an instance

I mean that's a terrible comparison given driving is far more easily accessible than the two mentioned examples. Plus how do you get the car to said race track?

Let's also not forget freight. 38-53% is ideally suited to switch to rail.

 

[al78]

Regardless of the benefits of fully autonomous cars, they are unlikely to have any significant positive impact on the traffic jams that many of our urban areas suffer from.

I disagree. A proportion of the inefficiency of the urban road network is because drivers cannot simultaneously move when traffic lights change to green (or more generally, cannot coordinate with each other to maximise efficiency). My home town is a classic case, the green phase on some junctions is no more than 10 seconds, and it can take nearly 5 seconds for the lead driver to wake up and move off, cutting the number of vehicles that can get through the lights by nearly half. This reduces the throughput of light controlled junctions. In fact, if all cars on the road were driverless, you could dispense with traffic lights entirely. Traffic lights are merely a blunt tool for different streams of traffic to communicate with each other. With driverless cars, each vehicle could communicate its intended path with every other vehicle in the vicinity and avoidance paths could be calculated and implemented on the fly. Another way driverless cars could reduce congestion is to always position themselves on the motorway equidistant from the vehicle in front and the vehicle behind, and maintain that distance. This prevents the irritation of traffic waves where everyone periodically grinds to a halt then the congestion clears with no apparent cause. It is explained here:

 

[Trainer2]

Driverless cars are the next transport game changer, railways in the majority of circumstances will be mothballed

 

[Envy123]

Driverless cars are the next transport game changer, railways in the majority of circumstances will be mothballed

They won’t decrease my rail usage. Going to London from my town is always faster by rail. Unless said cars will fly

 

[Bald Rick]

Let's also not forget freight. 38-53% is ideally suited to switch to rail.

Where do you get that number from?

 

[Tazi Hupefi]

Driverless cars are the next transport game changer, railways in the majority of circumstances will be mothballed

Maybe in your next life too!

 

[Annetts key]

I disagree. A proportion of the inefficiency of the urban road network is because drivers cannot simultaneously move when traffic lights change to green (or more generally, cannot coordinate with each other to maximise efficiency). My home town is a classic case, the green phase on some junctions is no more than 10 seconds, and it can take nearly 5 seconds for the lead driver to wake up and move off, cutting the number of vehicles that can get through the lights by nearly half. This reduces the throughput of light controlled junctions. In fact, if all cars on the road were driverless, you could dispense with traffic lights entirely. Traffic lights are merely a blunt tool for different streams of traffic to communicate with each other. With driverless cars, each vehicle could communicate its intended path with every other vehicle in the vicinity and avoidance paths could be calculated and implemented on the fly. Another way driverless cars could reduce congestion is to always position themselves on the motorway equidistant from the vehicle in front and the vehicle behind, and maintain that distance. This prevents the irritation of traffic waves where everyone periodically grinds to a halt then the congestion clears with no apparent cause. It is explained here:

Keep in mind that my answer was in the context of more people using automatous vehicles, the poster indicated that the railways would be in decline.

Yes, if ALL road vehicles were automatous, you would do away with traffic lights and various other junction layouts that are designed to slow down or otherwise restrict human driven vehicles. This, along with closer distance between vehicles and co-ordinated control would increase the number of vehicles in a given space. But I don’t think these changes would be enough to eliminate congestion where it is bad now. Let alone if there are no trains. Unless remote working catches on big time with employees only going into to office occasionally.

And as said above, while there are vehicles on the roads that have a human in control, well, you still need traffic lights and the various restricted junction layouts.

Oh, and by the way, in some parts of the country, drivers are already moving by the time the traffic lights turn green, and when the change to red, people often jump the lights. It’s not right, but it happens.

Anyway, back to trains. If/when the signalling system is changed to the highest level of ETCS/ERTMS, then this will enable the possibility of higher lines speeds and closer distances between trains, especially at slower speeds. Instead of the signalling system using fixed length sections (between signals), instead the distance can be reduced down to the service braking distance (plus a short safety margin). Also, the length of trains becomes much less important (from the signalling point of view), so where there are currently limits on the length of trains (due to the current signalling system), it may be possible to run longer trains. Combined, this will increase line capacity, thus in theory, in the long term, making trains more economic.

So, on a fast main line, if the track can be upgraded for say 155 MPH, then that’s the speed a suitable train could go. I don’t think it’s very likely that many automatous vehicles will be able to match that. Let alone offer the same capacity at that speed that a train would.

 

[SuperNova]

Where do you get that number from?

CITL rail freight forum

 

[HSTEd]

With driverless cars, each vehicle could communicate its intended path with every other vehicle in the vicinity and avoidance paths could be calculated and implemented on the fly.

You mean... moving block signalling?

If they can't get it to work on a mainline railway, good luck getting it to work on a road network which is inifnitely more complex.

 

[camflyer]

Well known that filling stations weren't too bothered about handing out those paper receipts, often they didn't even write anything on them, just tore one off the pad and handed it over (and if you were friendly, you could get a handful, or even an entire pad if you asked!).

Same with taxi drivers. They often couldn't be bothered filling in the details so you'd often get given a blank one. If you got a friendly one you'd get a wad of empty receipts which was very handy for when submitting expenses claims at work!

 

[Dr Hoo]

Let's also not forget freight. 38-53% is ideally suited to switch to rail.

Where do you get that number from?

CITL rail freight forum

These sort of figures can be inferred from published transport statistics relating to (say) HGV movements of consumer goods or bulk traffics over particular distances. These are flows that rail freight certainly 'ought to have a look at'.

There are, however, several important caveats:

- The movement does have to be part of a fairly thick 'flow'. One HGV on its own between A and B is unlikely to be suitable for rail. Thirty HGVs per day between a major food processing plant and a cluster of supermarket national or regional distribution centres (such as around Daventry) far more so.
- There need to be suitable terminals at either end of the trip. The non-rail-connected quarry or the major bottling plant in an essentially rural area will be hard to serve economically. Ditto the distribution centre next to a motorway junction but miles from a railway line.
- The general railway constraints of capacity, gauge, diversionary routes and so forth have to be favourable.

These and other factors tend to chip away at the 38-53% quite significantly. So not exactly "ideally suited".

 

[Bald Rick]

CITL rail freight forum

These sort of figures can be inferred from published transport statistics relating to (say) HGV movements of consumer goods or bulk traffics over particular distances. These are flows that rail freight certainly 'ought to have a look at'.

There are, however, several important caveats:

- The movement does have to be part of a fairly thick 'flow'. One HGV on its own between A and B is unlikely to be suitable for rail. Thirty HGVs per day between a major food processing plant and a cluster of supermarket national or regional distribution centres (such as around Daventry) far more so.
- There need to be suitable terminals at either end of the trip. The non-rail-connected quarry or the major bottling plant in an essentially rural area will be hard to serve economically. Ditto the distribution centre next to a motorway junction but miles from a railway line.
- The general railway constraints of capacity, gauge, diversionary routes and so forth have to be favourable.

These and other factors tend to chip away at the 38-53% quite significantly. So not exactly "ideally suited".

Indeed. There is absolutely no way 38-53% of Road freight is ‘ideally suited’ to rail.

To put it into context, maritime containers are ideally suited to rail, if the distance and flow rate is sufficient. Yet only 30% of containers leaving the port do so on rail.

 

[Starmill]

I think the really important thing to point out here is that rail as a mode within current technical capabilties actually is and will probably be able to remain competitive for lots of traffic in theory. The issue is that it spends lots of resources on markets it is technically very ill-suited to, especially rural routes through villages and small towns, but also some short distance trips on congested infrastructure shared with long-distance trains. Neither of these things is really sustainable.

It's also fairly clear that under investment of several decades means that rail isn't as competitive as it could be even in its prime market of long-distance travel between major cities, and medium distance commuting and leisure trips between large towns and the major cities. Freight strategy is also rather lacking.

Furthermore, some routes don't have a sustainable future regardless of how much investment they receive, such as the Conwy Valley line, and the sooner this difficult truth can be recognised the better for the rest of the industry really.

Perhaps what frustrates me more than anything else is that we're continuing to bake in new assets which are painfully unproductive and apparently ignorant of the main established strengths of the industry. Building stations at Reston or Soham as we speak are good examples. We're going to be lumbered collectively with these for decades, causing losses for nothing but short term political gain.

 

[quantinghome]

Driverless cars are the next transport game changer, railways in the majority of circumstances will be mothballed

They really aren't.

The technological challenges are immense. Just watch how Uber does over the next few years - their long term business model is reliant on removing drivers from the equation, yet it's looking like the technology won't be able to get there.

And even if somehow it does work, you've still got the basic problem of road capacity, which contrary to claims made by fans of driverless cars, will not increase. Junctions will still have conflicting moves. There is always going to be a reaction time, allowance for differential braking rates, safety margin. If anything this will increase the gap between cars rather than reduce it as driverless cars will be programmed to operate within safe margins, unlike their human counterparts!

 

[squizzler]

I think that motorcars could yet prove a dead end. The underlying horseless carriages which they are trying to automate are late Victorian technology that will be leapfrogged - in a literal and quite spectacular way - by the first adopters of personal jetpacks. Humans have for thousands of years dreamed of being able to fly at will, whereas the cultural idea that motorcars embody a form of personal freedom had to be manufactured over decades of the mid twentieth century. Given that ultimate form of personal freedom, I think we can see which the public will prefer.

We have been promised jetpacks since the fifties, if not before, but recent advances have brought them closer to reality. I saw one on Top Gear the other night, and it won a race against an electric racing car.

 

[Bletchleyite]

I think that motorcars could yet prove a dead end. The underlying horseless carriages which they are trying to automate are late Victorian technology that will be leapfrogged - in a literal and quite spectacular way - by the first adopters of personal jetpacks. Humans have for thousands of years dreamed of being able to fly at will, whereas the cultural idea that motorcars embody a form of personal freedom had to be manufactured over decades of the mid twentieth century. Given that ultimate form of personal freedom, I think we can see which the public will prefer.

We have been promised jetpacks since the fifties, if not before, but recent advances have brought them closer to reality. I saw one on Top Gear the other night, and it won a race against an electric racing car.

Personal jetpacks would be like a fast motorcycle. Personal drone-like e-helicopters would be the car replacement, but they'd have to be fully automatic and probably as a service.

 

[ComUtoR]

Driverless cars are the next transport game changer, railways in the majority of circumstances will be mothballed

There is something the train does that cars simlpy cannot compete with. It moves hundreds of people at the same time to the same destination. Driverless cars may be a game changer in terms of personal transport but mass transit is a different beast alltogether.

Take the morning commute into London. If each one of those people switched from the train to the car then the roads just become a sea of cars. The road infrastructure just won't cope with the capacity required. Also, the morning commute, although often a complete nightmare, has the benefit of being generally quite efficient. How long would a typical morning commute into London be by car ? When you arrive there will also be the issue of parking. How would you deal with the additional few thousand vehicles suddenly entering town centres ?

Even if you take leisure travel. Our seaside specials are full to the brim. Anyone who has taken their car down to the coast will know how much of a Drama Llama it can be. Imagine all those vehicles down the A303 !

TL/DR : Trains are cool.

 

[stuu]

In fact, if all cars on the road were driverless, you could dispense with traffic lights entirely.

Where do pedestrians and cyclists fit into this future technological utopia?

 

[Bletchleyite]

Where do pedestrians and cyclists fit into this future technological utopia?

Pedestrians and cyclists are best segregated from general traffic in urban areas.

 

[edwin_m]

Yes, if ALL road vehicles were automatous, you would do away with traffic lights and various other junction layouts that are designed to slow down or otherwise restrict human driven vehicles.

They would still have to cope with pedestrians, who are probably more unpredictable than most human drivers and certainly more vulnerable. Unless, as some have suggested, they are banned from the road except at controlled crossings, which in my view would be both morally wrong and unattainable. By the same logic cyclists would have to have segregated access to everywhere, and the resulting loss of roadspace would more than counteract any road capacity benefit from automation.

So if the concept is to work at all, it needs to be able to cope with unpredictable behaviour of other road users. It could be that autonomous mode is confined to motorways and similar roads that don't have these hazards, and where the likes of Tesla's Autopilot can nearly do the job now. This would still be a threat to railways.

 

[Bletchleyite]

There is something the train does that cars simlpy cannot compete with. It moves hundreds of people at the same time to the same destination. Driverless cars may be a game changer in terms of personal transport but mass transit is a different beast alltogether.

Take the morning commute into London. If each one of those people switched from the train to the car then the roads just become a sea of cars. The road infrastructure just won't cope with the capacity required. Also, the morning commute, although often a complete nightmare, has the benefit of being generally quite efficient. How long would a typical morning commute into London be by car ? When you arrive there will also be the issue of parking. How would you deal with the additional few thousand vehicles suddenly entering town centres ?

Even if you take leisure travel. Our seaside specials are full to the brim. Anyone who has taken their car down to the coast will know how much of a Drama Llama it can be. Imagine all those vehicles down the A303 !

TL/DR : Trains are cool.

What you probably want is a self-driving car to take you to the station for the train. Trains are good at mass transit.

They're a threat to country branch lines but not to the railway as a whole.