How could self driving vehicles effect future rail growth?

[Technologist]

Building new rail infrastructure is very expensive and especially so if it is designed for high speed trains. The UK is in a parlous economic situation with massive debt, whose servicing alone consumes vast portions of tax revenue.

Therefore, new rail lines should only be built if they can be demonstrated to have economic benefits, not just for political reasons.

Mods note - split from this thread

Restarting HS2a

Wonder how quickly the 3 Mayors can persuade Louise Haigh to restart HS2a to get it to Crewe ( using thier 3rd party / government funding) and reapply the safeguarding for HS2b

www.railforums.co.uk

Latter sentence is to a degree why the former is true.

The fact that we expect every line to be an economic net positive means that we are stuck doing similar things to what we have already done and thus have the economic data. Most of that data is also for rail that isn't anywhere near self funding and need to be justified on the basis of social or broader economic benefits.

This is not how the world's most successful companies are created or run, they generally don't make money for a long time (or at least don't make money on the early stages of their long term innovations) and they base their economics on general favourable trends (Moore's law, generally applied) which mean that as they scale they eventually become profitable and eventually incredibly profitable.

So Uber, Google, Facebook, Amazon etc all burnt money for years but they didn't do it wilfully. Those unprofitable ventures were all about building knowledge, capability and market position so that when the ventures scaled up they would be profitable.

This is why we need a top level grand plan integrated with the likely impacts of other technologies (this must not be done by railway people), I'd suggest that the long term vision should be that the railway becomes a walk up national high speed metro deeply integrated with last mile autonomous vehicles and where autonomous busses on dedicated routes duplicate or infill rail. This would be funded directly by capturing the economic benefit for the rail company itself which would then be used to keep developing the system and developing the technologies and business model, hence treasury and ministers are superfluous to the decision making. They instead become enablers on the planning, integration and economic benefits side

This would be under pinned by the fact that autonomous vehicles are making systems vastly more capable than the most advanced railway signalling commodity consumer electronics, batteries enable high performance trains to operate pretty much everywhere and modular and robotic construction can make building things surprisingly cheap to build.

The thinking then becomes where is the cheapest place to try all these things out rather than which line has the best cost to benefit ratios using my current paradigms. Find a relatively straight low usage line, chuck freight off it, partially electrify the easy bits of it and then run 300kph, tilting, lightweight BEMUs on it that use Starlink/ Amazon Leo/5G/GPS/5G/computer vision signalling systems to run the trains.

In other news the first coast to coast autonomous vehicle trip has taken place in the US. Yes you can apply all the caveats about US roads being easy and it being a one off and not verified by a third party but all those issues are going to be solved in much shorter time periods than rail can currently construct in.

2,700 miles: Tesla's first coast-to-coast autonomous drive completed

David Moss reported completing a journey across the United States using Tesla's Full Self-Driving (FSD).

interestingengineering.com

 

[stevieinselby]

In other news the first coast to coast autonomous vehicle trip has taken place in the US. Yes you can apply all the caveats about US roads being easy and it being a one off and not verified by a third party but all those issues are going to be solved in much shorter time periods than rail can currently construct in.

2,700 miles: Tesla's first coast-to-coast autonomous drive completed

David Moss reported completing a journey across the United States using Tesla's Full Self-Driving (FSD).

interestingengineering.com

If self-driving cars can be successfully deployed in the UK (and that's a very big "if") then any putative issues they might be a solution for (and again, that's a very sketchy case to make) are completely different from the very real issues that HS2 would be a very real solution to.

 

[Krokodil]

If self-driving cars can be successfully deployed in the UK (and that's a very big "if") then any putative issues they might be a solution for (and again, that's a very sketchy case to make) are completely different from the very real issues that HS2 would be a very real solution to.

Indeed. Self-driving cars are still cars. Electric cars are still cars. They still require huge amounts of land if they are not to cause gridlock.

 

[slowroad]

Indeed. Self-driving cars are still cars. Electric cars are still cars. They still require huge amounts of land if they are not to cause gridlock.

That’s why we need road user charges that vary by time, place, and type of vehicle. But still, most roads at most times have spare capacity.

A likely outcome is autonomous, often platooned HGVs at night, freeing up road space for cars at other times. (But agree this won’t solve the problem of inner city transport.)

 

[joieman]

That’s why we need road user charges that vary by time, place, and type of vehicle.

That seems like an excellent way to extort people who likely don't have the option of flexible working hours out of their hard-earned wages. That will never fly. Double taxation in the UK is already excessive.

 

[stevieinselby]

That seems like an excellent way to extort people who likely don't have the option of flexible working hours out of their hard-earned wages. That will never fly. Double taxation in the UK is already excessive.

In any vaguely sensible road charging plan, it would replace some existing charges – possibly VED, maybe a proportion of fuel duty – and so would see a lot of people's charges go down as well as other people's go up. Probably not quite revenue neutral, but total income would be roughly comparable to before the scheme while still bringing in a bit more money. People who generally drive small cars on quiet roads would pay very little, while those who drive honkin' big SUVs on heavily congested roads would pay more. That's no different to the principle of train fares being more expensive at busy times and cheaper when it's quiet, and not many people would object to that. If people were that unhappy about having to pay to drive in rush hour then it would encourage more people to reconsider whether they need ridiculously big cars, it would put more pressure on local authorities to improve public transport and active travel, and it would incentivise employers to allow flexible working where it could be accommodated. Ultimately, we need to take radical action to reduce road traffic, and especially to reduce road traffic where it is congested.

Do I think it will happen? No, there will be a million astroturfed campaigns on social media bewailing the betrayal of people's freedoms (despite it being abject nonsense), people who have no understanding of anything beyond what they see on Facebook, TikTok or the Daily Mail will be up in arms about it even if they stand to save significant sums of money, and no politician will have the guts to push through something that, if they did, could bring real benefits and people would quickly get used to.

 

[slowroad]

In any vaguely sensible road charging plan, it would replace some existing charges – possibly VED, maybe a proportion of fuel duty – and so would see a lot of people's charges go down as well as other people's go up. Probably not quite revenue neutral, but total income would be roughly comparable to before the scheme while still bringing in a bit more money. People who generally drive small cars on quiet roads would pay very little, while those who drive honkin' big SUVs on heavily congested roads would pay more. That's no different to the principle of train fares being more expensive at busy times and cheaper when it's quiet, and not many people would object to that. If people were that unhappy about having to pay to drive in rush hour then it would encourage more people to reconsider whether they need ridiculously big cars, it would put more pressure on local authorities to improve public transport and active travel, and it would incentivise employers to allow flexible working where it could be accommodated. Ultimately, we need to take radical action to reduce road traffic, and especially to reduce road traffic where it is congested.

Do I think it will happen? No, there will be a million astroturfed campaigns on social media bewailing the betrayal of people's freedoms (despite it being abject nonsense), people who have no understanding of anything beyond what they see on Facebook, TikTok or the Daily Mail will be up in arms about it even if they stand to save significant sums of money, and no politician will have the guts to push through something that, if they did, could bring real benefits and people would quickly get used to.

Agree completely. Most road users would benefit in net terms, particularly once time savings have taken into account. And rail subsidies could be cut, as part of the rationale would have disappeared.

 

[HSTEd]

I think Autonomous vehicles are conceptually an enormous threat to the railway. If they become widely available and comparatively inexpensive (and we have no reason to believe they will not, in time), they allow car levels of convenience to pretty much everyone outside of a few edge cases. It is the final triumph of the vision of Milton Keynes.

In that environment, railways have to somehow beat the hypothetical car journey in order to have a market at all.
The railway has to be fast, it has to be frequent, and it has to be cheap.

This is an enormous challenge on par with the first emergence of the mass market car, but I don't think the railway is doomed even in that situation.
Rail freight probably doesn't do very well when autonomous vehicles can remove a major cost of HGV road freight (staffing), and allow vehicles to operate at whatever time is most convenient without regard for drivers hours.

Roads would become heavily loaded 24/7, which might have some interesting noise implications but probably cuts congestion significantly.

On the other hand, social and political cover to abandon much of the railfreight market (all but heavy haul) would significantly reduce the subsidy levels on the railway and allow a refocussing on the passenger sector.

The railway has to begin preparing the ground for a case for major investment to justify its own existence in this environment. It needs to cut its cost base and be seen to be doing so to make the case for more money later. We also probably have to set out a long term vision to escape many of the fundamental limitations of the British railway network, especially in terms of train length, loading gauge and ancient and dodgy infrastructure.

Certainly a challenge on par with the run up to Beeching, if I'm honest - but the railway can weather this if it moves quickly and surely, and in the correct direction.

 

[NER1621]

Apologies, I’m a little confused. The thread title is “How could self driving vehicles effect future rail growth?” which I took to mean “how could SDV generate more growth for the railway?”

Given the trend of the discussion, which is mainly about negative impacts, should the title be amended to “How could self driving vehicles affect future rail growth?

 

[Sorcerer]

Here we go again. Sigh.

Autonomous and electric road vehicles will still have the same issues as petrol and diesel ones and will require the same amount of infrastructure footprint. They will never be able to match railways in their ability to move large amounts of goods and passengers over long distances, and if the invention of the road vehicle wasn't enough to kill off the railways then simply changing their means of movement will not do so either.

Tesla obviously has a vested interest in convincing people otherwise, but in reality the only way railways could meet their demise at the hands of road travel is if those with power and influence actively work towards it (such as holding back on investments and modernisation) rather than simply being an inherent inability to compete. Road travel has it's place, and that is not replacing railways.

 

[HSTEd]

Here we go again. Sigh.

Autonomous and electric road vehicles will still have the same issues as petrol and diesel ones and will require the same amount of infrastructure footprint.

Road vehicles already make up a huge portion of total traffic.
It would not take that much additional infrastructure to eat up rail's share too, especially if you exclude extremely specialist situations such as metros.

They will never be able to match railways in their ability to move large amounts of goods and passengers over long distances, and if the invention of the road vehicle wasn't enough to kill off the railways then simply changing their means of movement will not do so either.

The railway has a lot of operations that don't really fit into either of those categories though.
We have a lot of very lightly used railways maintained for social or political reasons, and those railways would be in serious trouble in an autonomous vehicle future.

Indeed, the rationale for most non-heavy haul railfreight looks shakey once driver shortages/pay are out of the picture on the road vehicle side. Especially as the autonomous vehicles would be able to run whenever is convenient in traffic and business terms.

 

[Oxfordblues]

I'm 73 so I don't expect to see self-driving vehicles on any scale in my lifetime. But just supposing they came along in any numbers we could expect some big changes. Just to take deep-sea intermodal: at London Gateway DP World is investing heavily with the intention of fully automating container handling and train loading. Similarly at inland terminals AI-controlled handling equipment could eventually eliminate manual unloading. And deliveries by self-driving trucks could further reduce labour costs. Ultimately the trains themselves might one day be self-driving (don't tell ASLEF!). After all the DLR happily operates without drivers, so the technology is already with us. In my dreams there could even be intermediate locations where containers would be automatically transferred between trains, opening-up many new traffic opportunities. All very "science fiction" of course!

 

[stevieinselby]

I think Autonomous vehicles are conceptually an enormous threat to the railway. If they become widely available and comparatively inexpensive (and we have no reason to believe they will not, in time), they allow car levels of convenience to pretty much everyone outside of a few edge cases. It is the final triumph of the vision of Milton Keynes.

In that environment, railways have to somehow beat the hypothetical car journey in order to have a market at all.
The railway has to be fast, it has to be frequent, and it has to be cheap.

This is an enormous challenge on par with the first emergence of the mass market car, but I don't think the railway is doomed even in that situation.
Rail freight probably doesn't do very well when autonomous vehicles can remove a major cost of HGV road freight (staffing), and allow vehicles to operate at whatever time is most convenient without regard for drivers hours.

Roads would become heavily loaded 24/7, which might have some interesting noise implications but probably cuts congestion significantly.

On the other hand, social and political cover to abandon much of the railfreight market (all but heavy haul) would significantly reduce the subsidy levels on the railway and allow a refocussing on the passenger sector.

The railway has to begin preparing the ground for a case for major investment to justify its own existence in this environment. It needs to cut its cost base and be seen to be doing so to make the case for more money later. We also probably have to set out a long term vision to escape many of the fundamental limitations of the British railway network, especially in terms of train length, loading gauge and ancient and dodgy infrastructure.

Certainly a challenge on par with the run up to Beeching, if I'm honest - but the railway can weather this if it moves quickly and surely, and in the correct direction.

If you're envisaging people buying self-driving cars then I don't see that making any appreciable difference to the railways – they will be more expensive than current cars, even allowing for the technology to become cheaper over time. They might in a small number of cases allow a household to make more use of one car – eg one person goes to work in the car then sends the car back home for others to use during the day, until it comes back to pick them up in the evening – but it would still incur running costs and would take time, it feels like quite an edge case where that would enable people to stop using the train so I can't see that as having any significant impact on rail use.

Or if it's just going to be self-driving taxis, where the cost of the technology may eventually drop to just below the cost of a driver's wages (but with little incentive for the vulture capitalists behind it to reduce it any further), then again it's unlikely to have a massive impact. Passengers will still face the challenge of availability and congestion at busy times (which, by definition, is when most people want to travel), robotaxis will be stuck in the same traffic jams that cars and taxis are now (except they'll be worse with robotaxis travelling between jobs or self-driving cars going off to park somewhere cheaper), and they won't go quantum in the rush hour, each one will still only be one and there will be a limit on how far the companies will go to meet peak-time demand when it results in hundreds of vehicles sitting idle and earning nothing for 18 hours a day. They won't be any faster than cars are now (for some drivers they might be slower, if they stick to speed limits!).

Most train journeys are either travelling into a major city, where congestion on the roads means that travelling by car or robotaxi is slow and therefore not convenient – or travelling long distances, where robotaxis at 70mph can't compete with trains at 100mph+, the only advantage over driving now is that you might not need quite as many stops at service stations (depending on whether your stops are determined by your tiredness or your bladder). Even on provincial services where neither of those apply, how many train passengers would say "I would give up the train and switch to taxis if they were 10% cheaper"?

There may be more impact on freight, if self-driving lorries are deemed to be not only viable but safe in transit – and I don't mean "they won't crash", I mean "they won't be susceptible to being waylaid by bandits who figure out how to stop an unaccompanied truck and steal or damage the cargo" – I know that's still a risk with a human driver, but I would have thought they would be less vulnerable and would have more options available to them.

What transport problem do self-driving cars actually solve?

 

[HSTEd]

If you're envisaging people buying self-driving cars then I don't see that making any appreciable difference to the railways – they will be more expensive than current cars, even allowing for the technology to become cheaper over time. They might in a small number of cases allow a household to make more use of one car – eg one person goes to work in the car then sends the car back home for others to use during the day, until it comes back to pick them up in the evening – but it would still incur running costs and would take time, it feels like quite an edge case where that would enable people to stop using the train so I can't see that as having any significant impact on rail use.

The cost of the car will only be marginally increased compared to normal cars, because fundamentally the only extra components will be electronics - and electronics have universally trended down in price over time.
Even loading the cars with LIDARs and the like is only going to add hundreds of pounds to its price when we are talking about the volumes involved here.

A self driving car will be able to drive you (or anyone or anything you want) from any location to any other location on demand. And it would be able to do it regardless of the physical condition of the person being carried, if someine is being carried at all.

A major factor in favour of the train is that driving is a time consuming activity that in a large number of cases is not enjoyable, and legally cannot be done if drunk. If I can work or play on my phone/computer in the car, what advantage does the train have left outside of the comparatively small number of cases where the journey is decisively faster end to end, allowing for transfers and waiting at stations?

Or if it's just going to be self-driving taxis, where the cost of the technology may eventually drop to just below the cost of a driver's wages (but with little incentive for the vulture capitalists behind it to reduce it any further), then again it's unlikely to have a massive impact.

The only way the "vulture capitalists" will be able to stop the price collapsing to the marginal cost is if they get absolutely everyone with the capability to build a driverless car to cooperate in an enormous cartel that sets the price of driverless taxi services and never sells a driverless car to a third party. Which includes various state owned enterprises in China and elsewhere. If any manufacturer breaks ranks, the cartel will disintegrate.

This is extremely unlikely, especially as it would reduce takeup of driverless cars and states will have clear incentives to encourage takeup.
Driverless cars would likely be much cheaper way to provide rural and socially necessary public transport over huge parts of the country than conventional buses etc.

Most train journeys are either travelling into a major city, where congestion on the roads means that travelling by car or robotaxi is slow and therefore not convenient – or travelling long distances, where robotaxis at 70mph can't compete with trains at 100mph+, the only advantage over driving now is that you might not need quite as many stops at service stations (depending on whether your stops are determined by your tiredness or your bladder). Even on provincial services where neither of those apply, how many train passengers would say "I would give up the train and switch to taxis if they were 10% cheaper"?

A very large fraction of rail journeys occur on a small portion of the system. 1 in 6 rail journeys on the Network Rail system occur on Crossrail alone!
Whilst technically rail will survive, the system will look like it got Serpell'd or worse, unless the railway acts iwth urgency to secure its position.

There may be more impact on freight, if self-driving lorries are deemed to be not only viable but safe in transit – and I don't mean "they won't crash", I mean "they won't be susceptible to being waylaid by bandits who figure out how to stop an unaccompanied truck and steal or damage the cargo" – I know that's still a risk with a human driver, but I would have thought they would be less vulnerable and would have more options available to them.

Thefts from moving vehicles are not really a serious problem.
I'd expect a net reduction in thefts because lorries won't end up parked in laybys overnight because of driver hours. They will proceed directly from their start point to their end point.

 

[stevieinselby]

A self driving car will be able to drive you (or anyone or anything you want) from any location to any other location on demand. And it would be able to do it regardless of the physical condition of the person being carried, if someine is being carried at all.

Taxis already do that!

Sure, you can't always get a taxi when you want it ... but that situation will be exactly the same with robotaxis because there will not be an infinite supply of them. Especially in rural areas, there might be a very limited deployment of robotaxis meaning that people could have a long wait and potentially a significant surcharge if the vehicle has had to cover a lot of dead mileage. You can hype it all you want, but fundamentally it's just a slightly specialised taxi, nothing more than that.

 

[HSTEd]

Taxis already do that!

Sure, you can't always get a taxi when you want it ... but that situation will be exactly the same with robotaxis because there will not be an infinite supply of them. Especially in rural areas, there might be a very limited deployment of robotaxis meaning that people could have a long wait and potentially a significant surcharge if the vehicle has had to cover a lot of dead mileage. You can hype it all you want, but fundamentally it's just a slightly specialised taxi, nothing more than that.

A robotaxi is a taxi with very low incremental running costs, especially if battery electric - which essentially all will be.

Labour costs mean that taxis are not a particularly affordable transport system, because they require a driver who has to be paid, sleep etc.

EDIT:

It's a taxi that has a cost per vehicle-mile comparable to a self driven car, it might even be lower if the technology achieves lower accident rates (and thus insurance rates).
So whilst its "just a taxi", its one with seismic potential impact.

 

[bspahh]

Taxis already do that!

Sure, you can't always get a taxi when you want it ... but that situation will be exactly the same with robotaxis because there will not be an infinite supply of them. Especially in rural areas, there might be a very limited deployment of robotaxis meaning that people could have a long wait and potentially a significant surcharge if the vehicle has had to cover a lot of dead mileage. You can hype it all you want, but fundamentally it's just a slightly specialised taxi, nothing more than that.

Robotaxis are going to be focussed on built up areas early on. Developing the software is a fixed cost. Once you have done that, the hardware for each extra car doesn't cost much. You will be able to park them close to each other, and they can drive themselves to park somewhere cheap if the parking spaces near a station are busy.

 

[Krokodil]

When Uber was introduced, people said "great, cheaper taxis means fewer issues with parking". What happened in reality in many American cities was that congestion increased because while there was less need to park, they instead generated a load of deadheading journeys.

The only solutions to traffic jams are viable alternatives to cars.

 

[778]

A major factor in favour of the train is that driving is a time consuming activity that in a large number of cases is not enjoyable, and legally cannot be done if drunk. If I can work or play on my phone/computer in the car, what advantage does the train have left outside of the comparatively small number of cases where the journey is decisively faster end to end, allowing for transfers and waiting at stations?

It may not be enjoyable a lot of the time (not a driver myself), but a lot of people do love driving and would not want to give up thier licences. I am not sure car enthuiasts would like autonomous road vehicles. If you did a poll asking people what tasks they would like to become automated, driving would be way down the list. They would much prefer an ironing robot or a vacuum cleaning robot over an autonomous road vehicle.

Whilst technically rail will survive, the system will look like it got Serpell'd or worse, unless the railway acts with urgency to secure its position.

Don't think it would get that bad even if self driving vehicles do take over. Long distance intercity routes and commuter lines will be ok, it is just the rural routes that will be at risk. I think at least 60-70% of the network would survive.

 

[HSTEd]

It may not be enjoyable a lot of the time (not a driver myself), but a lot of people do love driving and would not want to give up thier licences. I am not sure car enthuiasts would like autonomous road vehicles. If you did a poll asking people what tasks they would like to become automated, driving would be way down the list. They would much prefer an ironing robot or a vacuum cleaning robot over an autonomous road vehicle.

I am sure a lot of people do like driving on roads like Snake Pass and the like. But how many people enjoy driving down the same bland dual carriageway they've driven down 100 times before?

As you say, I doubt people would actively give up their licences or anything like that. The cost of keeping an old manually driven car around probably won't be too large, at least for two or three decades. By then the hobby will likely be comparable to horseriding in scope, I don't think it would be a major driver on transport policy. Of course assuming self driving cars do actually take off (I think it is more likely than not, but nothing is truly guaranteed before it happens).

Don't think it would get that bad even if self driving vehicles do take over. Long distance intercity routes and commuter lines will be ok, it is just the rural routes that will be at risk. I think at least 60-70% of the network would survive.

I think it would depend a lot on how the politics interacts with railway funding and the roads budget.
This is why my original post emphasised the political aspect, if the railway is seen to be taking positive and serious steps to addressing this challenge, the politicians will likely be looser with the purse and more can be saved.
I am doubtful we can save everything, but I think we could save most of it.

If the railway ignores this threat until it is upon them, we risk the industry being pressured into mass closures for want of economies, just like Beeching (the argument that more could have been saved with paytrain etc)

 

[Topological]

I am not sure I would want to give up my car. It took me a long time to learn to drive, but now I have I could not go back.

Robotaxis may be great. They may turn up the moment I decide to go out. They may be available for the whole day I have in mind. They may offer me the flexibility to change my plans en route. There could be lots of practical innovations like this that normal taxis do not offer. However, a taxi is a taxi.

People like to own things. I may be way off, but owning a car means a perceived achievement and a perceived freedom. No matter how much a communal solution addresses those, the psychological ownership factors remain.

Trains do an excellent job for what they do. They are a link to something powerful and beyond ownership. I think that is why people prefer trains to buses on the whole. The train is just a stronger thing. Planes beat trains for similar reasons. Some people might want to swap their train for a robotaxi, but I would not and I am sure many others would not.

In short, I do not see autonomous vehicles doing much to railway demand. Certainly nothing like the impact cars had.

 

[HSTEd]

I am not sure I would want to give up my car. It took me a long time to learn to drive, but now I have I could not go back.

Robotaxis may be great. They may turn up the moment I decide to go out. They may be available for the whole day I have in mind. They may offer me the flexibility to change my plans en route. There could be lots of practical innovations like this that normal taxis do not offer. However, a taxi is a taxi.

Well I doubt most people would give up their car, but they simply wouldn't have to drive it any more.

The robotaxi, as opposed to the personally owned driverless vehicle, would mostly be for people who don't have cars in the current society.

 

[Bald Rick]

Interesting discussion.

If fully autonomous road vehicles happen then it will affect a relatively small part of the rail market.

Most people choose rail over road because rail is better for them in their personal price / journey time / convenience equation. To switch that choice to road means something needs to change in that equation, and the simple fact that their car (or a car) can drive itself is unlikely to change it much, with the possible exception of avoiding parking charges.

There is a small proportion of the rail market who choose rail because they can’t (or don’t like to) drive. For them the equation is quite different. Assuming they are content with being driven by a ’robocar’ (and some people won’t be) then the cost part of their equation might shift fairly substantially, and this will undoubtedly move some people from rail to road. It will also shift people from buses and coaches to road too.

The segments of the rail passenger market most under threat will be parts of the rural/regional railway and a few secondary routes. The routes into major cities and long distance will be fine, as the road based alternatives there will ultimately be much the same as now in the price / journey time / convenience equation. Fortunately, the rural railway already loses a phenomenal amount of money, so the loss of income there won’t alter their economics that much. Although I suspect there will come a time when the almost complete absence of passenegers from some rural services does start to raise questions.

 

[slowroad]

Interesting discussion.

If fully autonomous road vehicles happen then it will affect a relatively small part of the rail market.

Most people choose rail over road because rail is better for them in their personal price / journey time / convenience equation. To switch that choice to road means something needs to change in that equation, and the simple fact that their car (or a car) can drive itself is unlikely to change it much, with the possible exception of avoiding parking charges.

There is a small proportion of the rail market who choose rail because they can’t (or don’t like to) drive. For them the equation is quite different. Assuming they are content with being driven by a ’robocar’ (and some people won’t be) then the cost part of their equation might shift fairly substantially, and this will undoubtedly move some people from rail to road. It will also shift people from buses and coaches to road too.

The segments of the rail passenger market most under threat will be parts of the rural/regional railway and a few secondary routes. The routes into major cities and long distance will be fine, as the road based alternatives there will ultimately be much the same as now in the price / journey time / convenience equation. Fortunately, the rural railway already loses a phenomenal amount of money, so the loss of income there won’t alter their economics that much. Although I suspect there will come a time when the almost complete absence of passenegers from some rural services does start to raise questions.

I think the long term effects are unknowable. Combined with hybrid working autonomous vehicles may lead to decentralisation of jobs, with big effects on modal choice. And for some routes, making HGV journeys autonomous and nocturnal could free up a lot is space for cars.

It will be interesting to see what happens to the far north lines, particularly in the context of (faster) electric buses for longer journeys.

 

[Topological]

Well I doubt most people would give up their car, but they simply wouldn't have to drive it any more.

The robotaxi, as opposed to the personally owned driverless vehicle, would mostly be for people who don't have cars in the current society.

I may be proven very wrong, but I am certain that we will not see a major modal shift to driverless vehicles owned or otherwise.

To create a market for robotaxis you need to replace taxis. I can see that happening, but not where the economics of the market lie. You save on a driver, but it is a big investment for someone to make and the alternative is simply not bothering to enter the market allowing your capital to go into any number of better schemes.

Then there are people who use cheaper forms of transport. Here you are not only paying the capital costs, but also talking about a market that does not pay as much currently. This is the first potential market that adds traffic versus the bus/train that used to run. Hence, in taking this market you risk devaluing the product.

Finally, there are people who do not use transport. These people may have underlying preference for road transport but not want to pay. Will they pay for the robotaxi? This segment also increases traffic and therefore devalues the product.

That then leaves the market that does not affect rail, the people who drive their own cars. Now that is tight market to enter because those people currently pay marginal cost, so you are trying to bring your marginal costs lower and then create a margin. Hard work without major scaling.

I see these self driving cars in much the same way as Dolby Surround Sound, or 3d television, both undoubtedly better than the existing product, but ultimately technologies that lesser systems will try to emulate. Where people will buy a sound bar that technologically produces surround sound, or a tv that gives more realism, they are still not getting the pure experience and so it will be with self driving cars. Parking assist etc., nice, full self-driving not adopted.

 

[Bald Rick]

It will be interesting to see what happens to the far north lines, particularly in the context of (faster) electric buses for longer journeys.

Yes I agree with that. No doubt the FNL is busy on some occasions, but it would be interesting to know why people actually use it. The last three people I know who used it did so simply to say they‘d done it. Few other passengers on board their trains, and this was in late spring. Also, none of the three paid! (railway staff). Anecdote is not evidence, obviously.

 

[172007]

I have said many times on simular threads.

True self driving cars will cause gridlock. Think about a family of 4. 2 x adults with a car then their 10 and 12 year old children. Currently there will be one or two cars on the drive. True driverless cars you will mean 4 cars, one each. And when the kids go around to their mates birthday party and 10 driverless cars all turn up on a street to park immagine the chaos. Immagine Schools with all the cars.

If driverless cars go down the Uber route of being taxi's then every journey will result in multiple more road journeys. Vehicle drives to pick your up, drives you tl your destination then goes to another pick up its next job so 2/3 journeys rather than one.

Grislock is my prediction or very slow journeys.

Raikways, especialy urban ones will be very busy as it wil be so much quicker to travel.

 

[Krokodil]

Rather a more pressing problem for rail:

https://www.reddit.com/r/trainwrecks/comments/1q75rs9/waymo_thinks_its_a_train/

The link is to a video of a Waymo car which took its passenger onto light rail track and started driving down it.

 

[stevieinselby]

If you did a poll asking people what tasks they would like to become automated, driving would be way down the list. They would much prefer an ironing robot or a vacuum cleaning robot over an autonomous road vehicle.

Quite. Whether or not you find driving a pleasure or a chore, if you're going from A to B in a car, you are stuck in the car for that time and there are a limited number of other things you can use that time for if you're not driving, so automating the driving part of it doesn't really free you up to do much else – I guess being a taxi service for teenage kids might be a different story if the only reason you are driving is to ferry them around – but yes, I would much rather automate actual chores that take up time where I could be doing something different.

 

[Sorcerer]

Rather a more pressing problem for rail:

https://www.reddit.com/r/trainwrecks/comments/1q75rs9/waymo_thinks_its_a_train/

The link is to a video of a Waymo car which took its passenger onto light rail track and started driving down it.

Yet again do we see the natural evolution in which all advancements in transit eventually leads us back to trains.