Windowless driver cabs?

[Andy25]

Seems to be a lot of sceptism about this but a lot of the technology is already around us. The main thing that will delay it being implemented is the slowness of the rail industry to adopt technology.

Too many people whistfully viewing driving as a human activity when we should be aiming to make computers do this for us (likely in a safer manor) so that humans can do more important tasks.

 

[underbank]

American roads are certainly easier to drive on, and busy cities are going to take much longer (harder to deal with the pedestrians/cyclists/etc), but actual driverless cars are very close to coming into operation across the pond.

Yes, in controlled/limited areas as I said. I have no doubt we may see the likes of a town or two, maybe Milton Keynes or the like being used as a test bed and being having widespread driverless cars, but it's just not going to happen any time soon in most towns/cities nor the countryside. Just because the technology may be just around the corner, doesn't mean that the practicalities will be feasibly dealt with until many new/updated versions, hence decades away.

 

[Andy25]

Yes, in controlled/limited areas as I said. I have no doubt we may see the likes of a town or two, maybe Milton Keynes or the like being used as a test bed and being having widespread driverless cars, but it's just not going to happen any time soon in most towns/cities nor the countryside. Just because the technology may be just around the corner, doesn't mean that the practicalities will be feasibly dealt with until many new/updated versions, hence decades away.

Won't be decades, gov is already having to put in legislation to make them road legal.

10 years I think, they've come a long way in the last 4 years and technology will only accelerate.

Only thing that will stop adoption will be people's lack of money to replace cars so will take time to replace conventional cars.

 

[Dave1987]

This attitude is what leads to economies becoming inefficient, and ultimately causing more unemployment in the long run. It's poisonous. People have to understand that some jobs will go away, the savings in one industry then lead to new jobs in other industries.

Trains put stagecoach drivers out of business at one point. E.g. a single train could've put 100 stagecoach drivers out of business (not counting those to take care of horses/maintenance/etc). It created new jobs in train operation and construction industries, more importantly it created new opportunities for business due to faster travel / easier goods transport / etc.

We no longer lament the loss of horse-drawn carriages and stagecoaches. Likewise, we won't lament the loss of train drivers jobs (just like not many in the real world lament the use of DOO). It's not quite the paradigm change that the introductions of trains saw, and it's going to take quite some time (e.g. it's easier/more cost effective on high frequency lines in a city, less so on infrequent rural lines) - but it will happen.

Anyhow, on the security topic: there's no reason trains should be hackable, first off because trains don't need to be connected to the internet. For now, everyone is connecting everything (including nuclear power stations) to the internet, but standards will evolve the more cracking goes on.

Like I said if you or the whole of society wants to plough billions and billions of pounds into trying to automate me out of a job I’m fine with that. I know how the real world works so I’m saying in my opinion driverless automated trains are fantasy stuff at the moment. I’m well aware how much money is being pumped into driverless tech and I’m well aware of how many serious crashes have happened in the Wild West that is the real world rather than the perfect world of a computer screen. I’m very happy to join the dole queue as well if it comes to it. Like I said I won’t be the first in the queue! Far from it. Sooner or later people won’t have enough money to pay for anything that the robots and AI have made....

BTW I’ve still yet to see the mass jobs being created by AI. Seems to be a handful here and there whereas I see the thousands of jobs it removing from the economy. So I hear this tech creating new jobs argument all the time but have yet to see any real evidence of it. Until I see the evidence I will continue with my “poisonous” attitude towards the tech industry.

 

[Peter Mugridge]

As to the use of screens in passenger cabins, you seem to be missing the point of windows entirely. I want natural light, not a pretty picture. If I was just going to go round looking at everything on a screen I might as well stay home and not travel at all. I have no doubt your giant claustrophic metal coffins would have great structural integrity, I just question whether they would have any passengers on them

This.

I certainly wouldn't want to travel unless I could see out of a proper window.

 

[6Gman]

Since "technology" hasn't yet worked out how to get rid of the manual signalling at Stockport (e.g.), and I could stand at the south end of Shrewsbury station on Tuesday and see adjacent platforms controlled by colour lights, upper quadrant semaphores and lower quadrant semaphores, you'll forgive me if I think it'll be a long time before we have a system to support trains that don't rely on drivers peering through windscreens.

 

[E_Reeves]

From the title I thought it meant windows without glass in

 

[takno]

Hate to break it to you, but driverless cars are actually extremely close:
https://arstechnica.com/cars/2018/0...es-of-testing-putting-it-far-ahead-of-rivals/

American roads are certainly easier to drive on, and busy cities are going to take much longer (harder to deal with the pedestrians/cyclists/etc), but actual driverless cars are very close to coming into operation across the pond.

// Edit: a more significant article: Lyft will actually begin operating self-driving cars in commercial service in the coming months:
https://www.cnbc.com/2017/11/07/way...-with-its-fleet-of-self-driving-minivans.html

Hate to break it to you, but beyond the spin none of the companies have really got that close to cracking the problem of driving safely in complex environments surrounded by other road users, even before Uber proved how far away they are from sorting the basics the other month. There's a huge amount of money at stake, and literally billions of reasons to pretend you're further ahead than you are, but the technology is easily 10 years away from being able to do the job in anything other than the most controlled environments.

There seems to be an assumption that trains are a simpler problem for some reason, and yet we've spent a lot of time and effort recently rather failing to prove that we don't need a human staff member from the other end of the train. I don't think there's an economic case to get rid of the driver even if the train could drive itself.

50-60 years ago we still had lift attendants pressing buttons for half a dozen people at a time. If you think that in the next 20 it's going to seem worth the cost of the crazily expensive technology and the risk of being the people whose robots killed hundreds just to save on the cost of a single salary, then you are clearly smoking something fantastic and I want in.

 

[James James]

Hate to break it to you, but beyond the spin none of the companies have really got that close to cracking the problem of driving safely in complex environments surrounded by other road users, even before Uber proved how far away they are from sorting the basics the other month. There's a huge amount of money at stake, and literally billions of reasons to pretend you're further ahead than you are, but the technology is easily 10 years away from being able to do the job in anything other than the most controlled environments.

There seems to be an assumption that trains are a simpler problem for some reason, and yet we've spent a lot of time and effort recently rather failing to prove that we don't need a human staff member from the other end of the train. I don't think there's an economic case to get rid of the driver even if the train could drive itself.

50-60 years ago we still had lift attendants pressing buttons for half a dozen people at a time. If you think that in the next 20 it's going to seem worth the cost of the crazily expensive technology and the risk of being the people whose robots killed hundreds just to save on the cost of a single salary, then you are clearly smoking something fantastic and I want in.

Uber aren't exactly representative of self-driving cars. Waymo are much better at this self driving thing - they've already driven millions of miles *on open road* not in controlled environments (which you'd see if you read the articles linked to upthread). They're starting commercial service *without a safety driver* *on open road* in a number of months (also in the articles upthread). No matter how strongly you deny it, that's commercial usage in an uncontrolled environment this year.

It is fair to point out that there's less money to be saved in automating trains, but trains are also a much simpler environment (no trains swerving onto your tracks, no hard-to-read traffic signs to parse, no dodgy road markings etc) - there's a lot less work needed for self-driving trains, and in fact they already exist, just in restricted environments. For more general you, all you'd need is to do is adapt the lidar and computer vision from self-driving cars to do obstacle detection (and ideally convert to ERTMS for the signalling). The hardware components are somewhat expensive right now (but being made cheaper by the self driving car development), but cheap compared to a train cost. There's more money to be made in developing this for self-driving cars, even if trains are easier, which is why we're seeing self-driving cars first.

Despite all that, people are already running truly driverless trains: https://www.popularmechanics.com/te...new-driverless-train-could-reinvent-shipping/

(It's worth remembering that technology development doesn't happen on a linear scale, past timeframes aren't all that useful in predicting how long existing technology can mature.)

 

[squizzler]

Hate to break it to you, but beyond the spin none of the companies have really got that close to cracking the problem of driving safely in complex environments surrounded by other road users, even before Uber proved how far away they are from sorting the basics the other month. There's a huge amount of money at stake, and literally billions of reasons to pretend you're further ahead than you are, but the technology is easily 10 years away from being able to do the job in anything other than the most controlled environments.

There seems to be an assumption that trains are a simpler problem for some reason, and yet we've spent a lot of time and effort recently rather failing to prove that we don't need a human staff member from the other end of the train. I don't think there's an economic case to get rid of the driver even if the train could drive itself.

50-60 years ago we still had lift attendants pressing buttons for half a dozen people at a time. If you think that in the next 20 it's going to seem worth the cost of the crazily expensive technology and the risk of being the people whose robots killed hundreds just to save on the cost of a single salary, then you are clearly smoking something fantastic and I want in.

You omit to say that, outside Silicon Valley fanboy circles, next to nobody in the real world is asking for this technology.

 

[takno]

You omit to say that, outside Silicon Valley fanboy circles, next to nobody in the real world is asking for this technology.

I don't think you can wait for people to ask for technology before you create it. One of the reasons for the hype around self-driving is an attempt to get people to want it, and think of markets and justify all the cost before you spend all the money. Personally I don't want a bunch of minicabs floating about the place getting in the way whether they have drivers in or not, and I see them as a largely see them as a half-baked solution to problem people are rather rudely and persistently failing to have.

 

[gallafent]

You omit to say that, outside Silicon Valley fanboy circles, next to nobody in the real world is asking for this technology.

… except, amongs others, the owners of road freight companies, and by extension the manufacturers of trucks, and the fleet operators and buyers. Once you have the choice between automation (with 24h running, and not having to pay a driver) and paying a driver who has to take breaks, ……… it all looks rather attractive economically.

https://www.ft.com/content/7686ea3e-e0dd-11e7-a0d4-0944c5f49e46 (apologies, the FT's Ts&Cs don't appear to permit copy-and-paste).

https://www.theguardian.com/technol...ving-trucks-automation-jobs-trucking-industry — a long article, a few snippets:

“And once the technology is proven, the incentive to adopt it will be powerful: in the US alone, large trucks are involved in about 350,000 crashes a year, resulting in nearly 4,000 fatalities.” — it's not just the economics that make the story compelling. In particular, letting the computer drive on the long, monotonous, relatively well-defined and controlled bits of road (where boredom and fatigue really hit hard for human drivers) just happens also to take out most of the hours the human driver currently works. That driver can now act as a local pilot at one or other end of the route, doing the “difficult” bit, close to home.

Much of the article is about the social and economic consequences rather than the tech, so I won't quote it here, but it's an interesting read.

https://www.wired.com/story/embark-self-driving-truck-deliveries/

“If you live in Southern California and you’ve ordered one of those fancy new smart refrigerators in the past few weeks, it may have hitched a ride to you on a robotruck.

“Since early October, autonomous trucks built and operated by the startup Embark have been hauling Frigidaire refrigerators 650 miles along the I-10 freeway, from a warehouse in El Paso, Texas, to a distribution center in Palm Springs, California. A human driver rides in the cab to monitor the computer chauffeur for now, but the ultimate goal of this (auto) pilot program is to dump the fleshbag and let the trucks rumble solo down the highway.”

If nobody were “asking for it” then nobody would be investing in it, or otherwise participating in the effort to make it happen. Still, it's well known that things which start off in silicon valley fanboy circles never make the big time, right?

There will be stops and starts as the technology develops, but the highway cruising capability is very nearly done now, and once it becomes sufficiently (significantly) less dangerous than a human driver, in that restricted context, I would hope it would become mandatory in short order. The retention of a “truck captain” to oversee the truck's operation might well be a sensible approach from a social and economic perspective, of course, …

 

[James James]

You omit to say that, outside Silicon Valley fanboy circles, next to nobody in the real world is asking for this technology.

This is actually a good point. Not that many people need this or are asking for it, and it completely ignores the issue that you need to move countries like the UK and US to public transport long term, because cars are inherently hugely inefficient.

But in the short term, road is king - and on-demand self-driving cars solve some of the issues of private car ownership, namely parking space, utilisation (most cars sit around doing nothing most of the time), etc.

There's a reason ride-sharing is huge in both the US and the UK - because dealing with your own car is a pain but the public transport is terrible. And that market is exactly where self-driving cars come in, because for the ride-sharing companies, self-driving cars are cheaper (and more flexible) than drivers. And if these cars prove to have lower accident rates (real-world trials seem to be producing good numbers so far - which isn't surprising, given most drivers are careless idiots), then it's not hard to see even car-owners switching to self-driving cars (that they happen to buy instead of paying for on-demand) - because insurances would bump up the rates for human drivers, and governments would legislate away the more dangerous human-driven cars.

It's worth remembering: there may be rare situations which a self-driving car can't handle (and more real-world usage will reduce the frequency of such situations), there are far more situations where humans cause accidents because they're drunk, tired, careless, plain stupid, or even malicious. I know, because I see such cases every single day.

And the more data cars get in the real world, the more data will be available for training self-driving trains - cheaply, and quickly. (Not only do we already have self-driving trains, but we want more of them because they allow more efficient infrastructure usage - so it's not even a question of self-driving trains, it's about trains that know when to self-stop, which is far simpler than driving a car.)

 

[furnessvale]

“If you live in Southern California and you’ve ordered one of those fancy new smart refrigerators in the past few weeks, it may have hitched a ride to you on a robotruck.

“Since early October, autonomous trucks built and operated by the startup Embark have been hauling Frigidaire refrigerators 650 miles along the I-10 freeway, from a warehouse in El Paso, Texas, to a distribution center in Palm Springs, California. A human driver rides in the cab to monitor the computer chauffeur for now, but the ultimate goal of this (auto) pilot program is to dump the fleshbag and let the trucks rumble solo down the highway.”

In a sensible world, 650 miles would be more than far enough to justify movement by rail freight.

 

[whhistle]

More technology that is relied upon to operate a service = more opportunity for things to go wrong.
Do you really want services cancelled because "the in cab camera systems have failed"?

No, but we already have trains cancelled because someone farted near the driver anyway! :P

It's already been suggested that computers are more reliable than humans when you take into account one human operator could control 4 or 5 trains at once. No spads unless the sensors fail of course - but I bet that wouldn't be as often as humans driving past red lights.

It's an interesting suggestion and one well worth exploring, but I think in reality terms, having trains without drivers all together will be the result. However, lessons could be learned through cameras instead of cab windows to prepare for no drivers at all. I mean, the driver could be sitting facing the side of the train if they like.

Of course though, drivers of this forum will be against it as anything that is a threat to their job is a no no, instead of working out that everything from video stores to supermarkets are adjusting and moving with technology. You have to find your place in the new world rather than stopping the inevitable.

It's not a pop at drivers BTW, just an observation.

Supermarket checkout workers have been under this threat for a long time. As many retire, move elsewhere, they replace the checkout with a self-checkout.

 

[gallafent]

In a sensible world, 650 miles would be more than far enough to justify movement by rail freight.

Absolutely agree. We're in a situation, though, in which infrastructure for trucks (driven by humans or computers) is already in place, and the same isn't true for trains. Same goes for air vs rail on some other routes — Kuala Lumpur to Singapore comes to mind — if the high speed line (or equivalent in-between infrastructure) doesn't get built (and SG/KL is now not planned to be), the alternative which needs infrastructure only at either end of the route (warehouses / airports) wins out over that which needs the not-yet-built infrastructure all the way in between as well.

 

[61653 HTAFC]

There's been talk of windowless planes for years, and there's far more of a reason to go down that road with planes due to the weight of windows/fittings able to go through pressure cycles. No sign of it happening though!

Windows on trains (in the saloons at least) don't need to be as thick or as heavily engineered, so there's less need to get rid of all that heavy glass. On the cabs, I can see the advantages in terms of collision resilience... and it would have avoided the whole 385 debacle I suppose!

 

[JohnMcL7]

According to wikipedia some of the class 20's have been fitted with a video camera to assist with forward visibility when not running cab first, is that the case? If so, are there any pictures of the system?

 

[Bletchleyite]

You omit to say that, outside Silicon Valley fanboy circles, next to nobody in the real world is asking for this technology.

Rarely does anyone ask for technology in that way. What they do want though is lower fares and someone in the passenger cabin to look after them. One way to achieve that is "guard"-only operation, such as has been in operation on the DLR since the 1990s.

 

[James James]

There's been talk of windowless planes for years, and there's far more of a reason to go down that road with planes due to the weight of windows/fittings able to go through pressure cycles. No sign of it happening though!

Windows on trains (in the saloons at least) don't need to be as thick or as heavily engineered, so there's less need to get rid of all that heavy glass. On the cabs, I can see the advantages in terms of collision resilience... and it would have avoided the whole 385 debacle I suppose!

Funny you should say that... because Emirates just started using fake windows, admittedly on the formerly-windowless inner first-class cabins:
https://www.theverge.com/2017/12/1/...lass-suites-virtual-windows-zero-gravity-seat

I imagine that will be a test-run for actually removing windows. With the exception of first class and the few window seated passengers, it's not like most of the passengers get to see much out the windows anyway...

 

[furnessvale]

I imagine that will be a test-run for actually removing windows. With the exception of first class and the few window seated passengers, it's not like most of the passengers get to see much out the windows anyway...

Especially with aircrew insisting that blinds are pulled down to watch some God awful film just as we are passing over the arctic wastes! Grrr!

 

[Roger100]

Rarely does anyone ask for technology in that way. What they do want though is lower fares and someone in the passenger cabin to look after them. One way to achieve that is "guard"-only operation, such as has been in operation on the DLR since the 1990s.

DLR driverless operation was possible because it is a very small railway with very slow trains, all doing much the same thing. The technology required to make it work was relatively small scale and thus not too expensive. Now to scale this up to manage 125+ mph trains over thousands of miles of complex track is another matter. Also the entire fleet of traction would need to be replaced or converted. Given the scale back of electrification of the GWML and other lines due to costs, I think full automation is likely to be kicked into the long grass for decades.

Road vehicle automation adds extra layers of complexity. On a railway the signaller instructs the driver, and the signaller has an overview of the relatively low number of vehicles on the tracks. The current aim of road vehicle automation is that each vehicle makes its own decisions. When I drive it can involve negotiation with other drivers - by eye contact, nods, gestures, light flashes etc. It also involves courtesy, like letting other drivers onto a busy main road from a side road. When parking at a large event there are often humans directing drivers to parking spots. Duplication of this process by some other communication would seem to be necessary. This seems to be a weak point - I have a nice fast fibre net connection now, but 25 years ago I would not have believed that there would be millions of computers hijacked into botnets sending out continuous spam, viruses and ransomware.

 

[Killingworth]

This.

I certainly wouldn't want to travel unless I could see out of a proper window.

Millions travel on underground lines where there's nothing to see out the window except tunnel walls and glimpses of stations between closely packed fellow travellers! Surely that's where technology can be best tested for windowless trains.

It takes decades for such things to happen, but imagine moving panoramas of surface London keeping pace with progress underground. Ah, they'd probably not be seen for the crowds of fellow travellers. Pictures of fish swimming past while in Eurostar..... Glaciers when in the tunnel on the way up the Jungfraubahn.....

A train controlled from a position that's not in the sharp end makes sense. Safety from objects dropped from bridges, or a major crash come to mind. Removing front windows and reinforcing the front of trains might let them pass more easily through tunnels festooned with winter icicles.

If forward looking cameras would make it easier to allow through corridor links between all units within a train that alone would be a major benefit. Seeing pairs of 140, 150 and 185 units operating locally with the crews unable to move between them seems nonsensical. Especially when 153, 156 and 158s don't have that problem. (Current Northern marshalling of 150/2s that can interconnect to other units that can't is just another aspect of their present crisis.)

 

[gallafent]

DLR driverless operation was possible because it is a very small railway with very slow trains, all doing much the same thing. The technology required to make it work was relatively small scale and thus not too expensive. Now to scale this up to manage 125+ mph trains over thousands of miles of complex track is another matter.

I reckon that's a tough job, yes. Might take around thirty years. Maybe even a little more.

https://www.theverge.com/2017/6/16/15815666/driverless-tgv-high-speed-trains-france-sncf

“France’s national railway operator aims to have autonomous high-speed TGV trains running by 2023, according to a report from FranceInfo. The train operator, known by the French acronym SNCF, will begin testing a prototype of its so-called “drone train” in 2019, FranceInfo reports.

“The new TGV train would be equipped with sensors that would allow it to detect obstacles and automatically brake, if necessary. The TGV, which reaches speeds of nearly 200mph, could be remotely piloted, though conductors will remain onboard in the short term to handle emergencies or unexpected events. SNCF President Guillaume Pepy tells Le Figaro that if the project is successful, SNCF would be the first operator in the world to run automated high-speed trains.”

[…]

— and yes, I expect that will be on the LGVs only, initially, with a human driver taking over when the train is on the classic network, including entering and leaving shared classic/LGV stations.

Also the entire fleet of traction would need to be replaced or converted. Given the scale back of electrification of the GWML and other lines due to costs, I think full automation is likely to be kicked into the long grass for decades.

You're might well be right, in the UK, sadly. The much-needed capacity enhancements that it would unlock without (even more expensive) new lines would be a boon. That's except for the key bits of infrastructure where it's already being deployed, of course (Thameslink core, and Crossrail).

Road vehicle automation adds extra layers of complexity. On a railway the signaller instructs the driver, and the signaller has an overview of the relatively low number of vehicles on the tracks. The current aim of road vehicle automation is that each vehicle makes its own decisions. When I drive it can involve negotiation with other drivers - by eye contact, nods, gestures, light flashes etc. It also involves courtesy, like letting other drivers onto a busy main road from a side road. When parking at a large event there are often humans directing drivers to parking spots. Duplication of this process by some other communication would seem to be necessary.

I think I disagree with the highlit sentence. Even now, satellite navigation systems use traffic information which is aggregated from data being collected from other such systems in order to plan the best route, one of the first forms of collaboration between vehicles — routeing choices are already made using shared knowledge. Further on from that, the technology is already being developed for autonomous vehicles to communicate in a much richer way than just eye contact, gestures, etc., instead transmitting either via centralised services (already happening), or over local mesh networks (possibly in prototype now but certainly not ready yet … so-called 5G short range high bandwidth communication systems are useful for that.

First we'd expect to see platooning (groups of vehicles known to each other) working together, but subsequently communication between any vehicle and another (“I'm four cars ahead of you and have just slammed the brakes on because a buffalo ran out in front of me, probably a good plan to act accordingly”, together with ~10cm accuracy location, speed, current road conditions (to allow correct brake application) data, context about the location of the rest of the herd of buffalo and their expected trajectories, amongst other things …), all in very small fractions of a second, would also happen, and really add to the capabilities of the road/vehicle system as a whole.

This seems to be a weak point - I have a nice fast fibre net connection now, but 25 years ago I would not have believed that there would be millions of computers hijacked into botnets sending out continuous spam, viruses and ransomware.

Security of these systems will indeed be paramount. The concept of “trust” and validating that a particular participant in a “conversation” is being truthful and “playing for the team” in these interaction scenarios, different vehicles having different views of the best set of actions for each member of the conversation to undertake (with asymmetric information, different decision algorithms and knowledge of each others' capabilities, and so on) ……… yes. And even worse, the possibility of actual hijacking of a given autonomous node's identity. Plenty of interesting work to be done.

 

[TwistedMentat]

I think people have jumped from my idea of augmenting the driver to a full on replcaement. My original post had nothing to do with driverless trains. Rather using technology to improve the safety and ease the workload of drivers.

Driverless or remotely operated trains are a completely different question.

I'm talking about using an array of cameras and on train computers to do stuff like identify and highlight signals, trackside obstructions, reduce glare, display things that would otherwise require looking down at the dashboard. And then taking advantage of modern high resolution monitors to move the driver into a safer or more efficient position.

And on the concept of windowless carriages I think people here tend to forget they're part of an edge case. Look at how many people around the world are used to sitting in buses with view disrupting viynals over the windows. How passengers in widebody airplanes happily sit in the middle set of seats with now view out the small windows. How passengers catch underground trains where the only real time you get a view is out onto a station platform. In spite of what people as a whole say, they're perfectly fine not having a view outside most of the time.

 

[takno]

I think people have jumped from my idea of augmenting the driver to a full on replcaement. My original post had nothing to do with driverless trains. Rather using technology to improve the safety and ease the workload of drivers.

Driverless or remotely operated trains are a completely different question.

I'm talking about using an array of cameras and on train computers to do stuff like identify and highlight signals, trackside obstructions, reduce glare, display things that would otherwise require looking down at the dashboard. And then taking advantage of modern high resolution monitors to move the driver into a safer or more efficient position.

And on the concept of windowless carriages I think people here tend to forget they're part of an edge case. Look at how many people around the world are used to sitting in buses with view disrupting viynals over the windows. How passengers in widebody airplanes happily sit in the middle set of seats with now view out the small windows. How passengers catch underground trains where the only real time you get a view is out onto a station platform. In spite of what people as a whole say, they're perfectly fine not having a view outside most of the time.

I hate buses with vinyls as well, but they let in plenty of light, and as with the underground trains you are rarely on them for very long anyway. Even the middle seat in an aeroplane gets a reasonable amount of natural light. That's physiologically and psychologically important to a lot of people even if the view is a bit rubbish.

The reasons you might want to do it on aeroplanes are due to the relatively huge cost of replacing very lightweight bodyshell materials with very solid glass and then firing it it upwards. In the case of train windows the cost difference of pushing them along compared to relatively heavy bodyshell materials is pretty negligible. You also want to mess around with peoples circadian rhythms on planes anyway if you are taking them across a couple of timezones, so there may actually be benefits to removing windows which certainly don't apply in trains.

 

[driver_m]

Are these metal windowless boxes going to cart robots about, because there's only going to be those who can use them. Humans won't have any money to do anything.

 

[birchesgreen]

The problem with windowless trains (or any other transport) is that you just know it won't only be a virtual representation of what you could see out of glass but plenty of adverts.

 

[underbank]

Underground is a special scenario as most journeys will be very short and for those who don't like them, there are alternative options such as overground and buses etc. I think most people can cope with 10/15 minutes on a tube (or whatever the average journey time is), but fewer will cope with 2-3 hours of it or longer.

For many travellers, it's the fact you get a view from the train that they choose the train in the first place, hence the complaints re lack of alignment between seats & windows and smaller windows. I think the railways would lose a lot of business (especially longer distance) if they ever made "the view" even worse as it would just drive people to air travel instead.

 

[birchesgreen]

55% of the London Underground is above ground anyway