I suspect this is the result of some civil servant being asked to find things they can do that will (a) cut the cost of trains, in principle but (b) not be offensive in the eyes of the Tories, which leaves a pretty small solution space. Hard to imagine this making any measurable difference, but it's at least politically inoffensive...
However, it's very much the train cost equivalent of telling people to use paper straws because you don't want to have the "you drive too much and eat too much meat" conversation on CO2 emissions...
I don't recall ever being on a train where the WiFi worked. If I need internet access while on the train I just tether to my phone. This usually works fine until you're in rural areas or in a tunnel. I find Vodafone gives best coverage outside of cities and towns, 3 can be awesome if you get lucky.
I once streamed a live F1 race while on the train from Edinburgh to London. There obviously were a few glitches but it really was good enough.
> According to a Transport Focus report cited by the DfT, a survey in December 2022 showed wifi on trains was a lower priority for passengers than other features – although those included such essentials as value-for-money fares, reliability, punctuality and personal security.
This is some really twisted logic, obviously people are going to prioritise punctuality and reliability before wi-fi if you can't even get your trains running on time. This doesn't mean people don't want wi-fi on the train. Have they really come to the conclusion that removing wi-fi is going to free up money to fix punctuality?
If the money would actually fix punctuality I suspect many would hold this as the correct priority -- I'm just not remotely convinced that wifi money would do that. If delays were really happening for want of such a pittance, they wouldn't be a problem in the first place.
> Improved internet and mobile connectivity on trains enhances rail’s
competitive advantage over other modes of travel and could also be instrumental in delivering the type of real-time information that passengers
want.
Monumentally stupid. In the post pandemic wfh Era, the way for trains to be viable is to offer amenities which let you "work from train". People still like face to face over zoom, but not enough to overcome the friction of commuting to a meeting place. The only way to reduce that friction is to turn commute time into viable productive time rather than a time sink.
At least some employers consider it feasible to work from a train, especially a long-distance train, and count it as hours worked. It's also useful when travelling to a meeting.
> A DfT spokesperson said: "Our railways are currently not financially sustainable, and it is unfair to continue asking taxpayers to foot the bill, which is why reform of all aspects of the railways is essential.
Well, yeah. Public transport never going to be as sustainable as private transport, that's why the government is funding it in the first place. It's partially to provide tax payers with a service, and partially to offer an alternative for those who can't afford other options.
Whoever put punctuality, personal safety, and train Wi-Fi in the same part of the questionnaire was setting the research up for killing the Wi-Fi. Obviously people prefer not being stabbed on the train above Wi-Fi and "arriving at the destination on time" is the main purpose of the whole field.
"We want commute to be worse and more expensive" is a bizarre take to have as the department of transport.
> Well, yeah. Public transport never going to be as sustainable as private transport, that's why the government is funding it in the first place.
Is private transport (i.e., the construction of roads) even financially sustainable? I did a quick search and it appears in the UK the amount brought in from vehicle tax each year is less than the amount spent on roads (https://www.gov.uk/government/statistical-data-sets/rde01-ro...).
No. But also very much yes, if you're talking about the economy as a whole.
Transport is massively profitable for the economy but not for the transport companies themselves, they simply can't capture enough of the value generated to turn a profit. If they try then the level of travel disproportionately goes down to make them lose even more money. It's why most/all of a public transport needs to publicly funded in some way and saving a penny or two per ticket on WiFi will cost far more to the economy than it saves.
There is one notable exception. Japan's train system is profitable depending on how you look at it. The trains themselves are a "loss leader" in that they actually lose money per train ticket, but the company that runs the trains also owns the terminals (lots of retail space) and the commercial real estate around them, which is where the money is made. They are in effect a commercial real-estate holding company that's doing transport on the side to beef up their portfolio.
"The DfT is also considering the cost of replacing or upgrading some on-train wifi equipment installed in the middle of the last decade, and whether passengers on shorter journeys use their own mobile phones or data rather than connecting to an operator’s wifi."
Well, based on my experience dealing with in-train coverage both working at a telco and as a traveler, if their hardware was deployed nearly ten years ago, then it is quite likely that their passengers _are_ using their own data plans instead of it.
In London you can rely on good coverage _outside_ the train, and I suspect (having visited 3 years ago) that the hardware is OK. The real trouble is when you do suburban/intercity trips, often through routes that have less coverage than motorways (which are prioritised by operators).
Why would motorways be prioritized? Maybe streaming music is more common than I think, but rail passengers have a lot more attention available to dedicate to a phone screen.
At least in Britain, a first class train ticket is generally considered fairly luxurious for longer trips -- mostly used by people travelling for business. Take a taxi to the station, the train to the required city, and another taxi to the destination.
Manchester to London tomorrow morning is £184 in standard class, £255 in first class. People/businesses who can afford that are choosing it because it's better than driving.
(Both tickets are significantly cheaper if booked in advance.)
I any case, there's voice service over almost the entire country. It's the fast data coverage that can be a little patchy on some more remote bits of railway lines.
> Our railways are currently not financially sustainable
I have raised this issue before, let me rehash it. Railways is probably the worst mode of travel in the history of mankind. I have not found any examples of large scale, passenger railway that is a sustainable business.
But people continue to hype it up and it's always "Let's spend more billions of other people's stolen money into this, it will definitely work out then!". It is never "Let me invest billions of my own money into this railways system."
If something has never worked, it is unlikely to start working now, unless you change the economics fundamentally. I haven't seen that in any system. The wastage is immense, and the money is better spent on things that actually provide more value for less cost.
To conclude, they should remove the entire railways instead of removing wifi from the trains.
Railways are an excellent mode of travel. They are significantly more efficient than cars in all ways that matter: cost, space, energy.
Nobody looks at the billions spent on roads and thinks "ooh these are unprofitable let's get rid of them", so why do people/governments constantly do that for rail?
The reasoning people use to justify the spending on roads is that the economic value people and goods transported on said roads outweighs the costs of building and maintaining them. Trains are more efficent than cars on this simply bu economics of scale: one long tube with metal wheels carrying 2000 people into the city moves a lot more than a freeway of the same area.
> They are significantly more efficient than cars in all ways that matter: cost, space, energy.
Then why are railways always unprofitable while car companies are worth billions in the free market?
> Nobody looks at the billions spent on roads and thinks "ooh these are unprofitable let's get rid of them", so why do people/governments constantly do that for rail?
Because roads are profitable and make sense. Rail does not.
Also, railways are at least 100x more expensive than roads. And the entire infrastructure has to be subsidised. With roadways, you just build the surface, vehicles are wholly privately owned. That means trillions less spent from the taxpayer's pocket.
The topic is British railways, where the trains are constructed, owned and operated by private companies. Railway vehicle manufacturing companies are presumably profitable, as they're all private. (At least the ones relevant to the UK.)
The track, electricity supply, signalling system etc in Britain is owned and maintained (and extended) by a public company, although it contracts out a lot of the maintenance work.
I am curious what you think a fair price for driving a car on the motorway from, say, London to Leicester (exactly 100 miles) should cost. Currently it's free, although fuel is taxed, perhaps around £5 for that journey.
Roads are massively subsidized. I suspect more than railways. If roads were tolled at the rate to maintain each segment then sure I would buy your argument. There is no way the maintenance of all of the roads is born according to actual use. It is done by also taxing those who do not use them.
First, there are other things to optimise for other than financial profitability - like travel time, emissions, comfort, land use efficiency, capacity, etc. (on all of those trains beat any alternatives in the "more than walking or short bike/car trip" to "less than where a plane makes sense (it varies a bit on geography, but I'd say ~4-5 hours by train)" category).
Second, there are profitable passenger railways - a very famous example is Japan's high speed rail system.
> Second, there are profitable passenger railways - a very famous example is Japan's high speed rail system.
No, the only reason Shinkansen exists is because of government subsidies, both direct and via interest-free "loans" without any repayment terms.
> First, there are other things to optimise for other than financial profitability
You don't have to optimise for financial profitability. If the product or service you provide is valuable to people, it will automatically bring in profits. If not, then you either go out of business, or subsist on continuous subsidies, e.g. every single passenger rail system on the planet.
> You don't have to optimise for financial profitability. If the product or service you provide is valuable to people, it will automatically bring in profits.
Nope. Railways are very capital intensive, thus by definition require a long time to reimburse those initial costs. So if companies never invest in any further infrastructure, at some point they'll become profitable and remain so (see: all american cargo rail companies that haven't done anything but the bare minimum infrastructure wise for decades). But demand isn't static, therefore to actually serve the needs of the population the infrastructure needs to evolve - more routes, faster ones, newer and better trains, etc. which requires constant capital. Also, providing a valuable service doesn't mean that it can be directly financially profitable, and this is a perfect example - the indirect benefits for the economy, environment and people (because more people can access more places for work, and emissions can be drastically reduced, etc.) can and usually drastically outweigh the direct financial cost of running the route.
> No, the only reason Shinkansen exists is because of government subsidies, both direct and via interest-free "loans" without any repayment terms
Like literally any public infrastructure ever? That's literally one of the major points of a government, to build infrastructure everyone needs and will use. Roads, bridges, airports, railroads are capital intensive and rarely if ever profitable on their own in the first decades of their use. Unless you're advocating for personal teleportation devices for everyone, I'm not even sure what your argument is supposed to be; but then again, you said "stolen momey" when refering to taxes, which makes you unqualified to weigh in on any topic around infrastructure, urbanism, government spending or organisation, etc.; anyone with such a misunderstanding simply does not know enough to be a useful participant in a discussion.
> So you're saying passenger rail can never provide enough value to people to justify the constant influx of capital? Cool, thanks for making my point.
Can not, or more likely should not receive enough money from consumers to make for the constant influx of capital needed for it to remain as useful as possible. The main goal is to be as useful and practical as possible, not to make the most money possible - a fully business class with a spa and golden jacuzzi train might be profitable (like the Orient Express was as a luxury thing) but again, the goal of passenger rail is to be accessible and practical
to most people. If your only way of measuring productivity is financial profit, yeah, I see you can't understand that.
> You literally said that Japanese high speed rail is profitable. Why change tracks (pun unintended) when I point out the inaccuracy?
Because you elaborated on your definition of profitability (can't have received any money at all from a government, ever), which no infrastructure project from the past few decades really meets. In my view the Japanese rail system is profitable after some initial well needed help; and again, if you apply your own (bad) measurement stick to any infrastructure, you'll find that nothing is really profitable. As for your argument that rail is the least profitable, I'm fairly sure if you do the math you'll see that Japanese rail has been less unprofitable (has needed less subsidies and bailouts) than Japanese airlines over the same time.
> Yes, and out of those only railroads are the one which are never profitable, even after 10 decades of constant technological advancement.
Airlines literally need bailouts on a what, decade schedule? But sure, railroads are "never profitable" because they get subsidies and need continued improvement to get better.
> I'm not advocating for anything except not wasting taxpayer money on inefficient, unsustainable transportation.
How is rail not sustainable? It's the cleanest and environment friendly method of transportation by far, which is very sustainable.
> Classic cope when you can't attack the argument, just resort to insults.
I am attacking your argument, but that doesn't mean I can't dismiss your entire position as one of ignorance (which, again, doesn't mean I'm not trying to explain to you why you're wrong). I cannot take libertarians seriously, only someone profoundly privileged or ignorant can have such views on the world.
>No, the only reason Shinkansen exists is because of government subsidies, both direct and via interest-free "loans" without any repayment terms.
Can't you say that for many "successful" car companies too? They're only around because of government bailouts and because of indirect government subsidies like road construction.
Technically, good and usable Wi-Fi on the trains makes sense - instead of having hundreds of phones all transmitting and interfering with each other, you have a handful of modems on either side of the train (with external antennas) providing the backhaul for the on-board Wi-Fi.
It should even be a win for the mobile providers too as it uses the spectrum more efficiently since less is wasted on interference.
> This is in Europe, mobile data coverage is practically 100%, isn't it?
Not really. In tunnels and remote areas (I mean middle of nowhere train tracks with fields, hills, cows as the only things within a multi-km radius) coverage can be extremely spotty. Even on very highly used high speed routes like Paris-Lyon or commuter routes like Paris - Melun there are blind spots (and big ones depending on the provider).
Based on personal experience using wifi in many metropolitan areas that provide it, it has been junk everywhere. I always use my data instead. If it was up to me I’d absolutely put that first on the chopping block for cost savings.
Either make it actually good and usable or discontinue it. As it stands, most "public Wi-Fi" projects I see around require too much effort to be actually usable. I bet the only reason we have Wi-Fi on the trains to begin with is someone's project to appear "modern" and skim some government money rather than provide a valuable service people can take for granted and actually start using and relying on. None of the people who worked on this actually use it day-to-day otherwise they'd realize how shit it is.
Public Wi-Fi "login" is absolutely awful. It relies on captive portals which 1) provide no security (the traffic is still unencrypted and you can take over someone's session by spoofing their MAC address, making the whole "auth" pointless), 2) don't actually fulfil any law-enforcement objective (malicious users will just provide fake details on the signup page anyway) and 3) break all kinds of applications and ultimately ruins the user experience to a stage where people just don't bother and tether to their mobile instead.
The solution is to either amend the law to make the whole "login" thing unnecessary (criminals will "login" with fake details anyway), make it open to everyone (so connecting immediately gives you unrestricted internet access with no captive portal, thus devices can auto-connect) and use some per-MAC-address and maybe per-domain traffic shaping to throttle down heavy users or heavy services (video streaming, etc).
Alternatively, if "login" or payment is needed, use WPA2/3-Enterprise (with EAP) which allows usernames and passwords (so the OS can handle the auth instead of a captive portal) with a single (potentially government- or telco-consortium-operated) login that works across all public networks so Wi-Fi actually works seamlessly and effortlessly. The Virgin Media Wi-Fi in the London Underground actually works similarly where users or any major carrier can seamlessly connect via EAP-SIM/EAP-AKA, never see a login page, and their device automatically roams on and off the network as needed.
Wait, there's a law requiring captive portals in the UK? Is that standard practice or unique to the UK? They are basically the worst part of public wifi, so that's a pretty unfortunate requirement
I can't remember to which trains this applies, it may well not even be the UK, but there are some railway carriages which unintentionally block 4G and 5G signals. When they were designed in the 1990s or so, it was a side effect of blocking UV or something like that.
In that case, adding WiFi inside the train, with the antenna outside the accidental Faraday cage, is worthwhile.
I have never been in a situation where my own mobile data didn’t work and the onboard Wifi did work. (as a seasoned commuter on TPE and LNER services).
If it’s one less thing to distract the TOCs from you know, operating trains, then I say get rid of the Wifi.
Contract out the provision of new wifi gear to private companies. Given IOT economies of scale this should diminish to an almost negligible cost. Piggyback security cameras and possibly entertainment screens on the same service. Could be ad supported to pay for itself and embarrass passengers by showing them ads for the things they just Googled.
50 comments
[ 4.8 ms ] story [ 95.5 ms ] threadHowever, it's very much the train cost equivalent of telling people to use paper straws because you don't want to have the "you drive too much and eat too much meat" conversation on CO2 emissions...
I once streamed a live F1 race while on the train from Edinburgh to London. There obviously were a few glitches but it really was good enough.
This is some really twisted logic, obviously people are going to prioritise punctuality and reliability before wi-fi if you can't even get your trains running on time. This doesn't mean people don't want wi-fi on the train. Have they really come to the conclusion that removing wi-fi is going to free up money to fix punctuality?
> Improved internet and mobile connectivity on trains enhances rail’s competitive advantage over other modes of travel and could also be instrumental in delivering the type of real-time information that passengers want.
Well, yeah. Public transport never going to be as sustainable as private transport, that's why the government is funding it in the first place. It's partially to provide tax payers with a service, and partially to offer an alternative for those who can't afford other options.
Whoever put punctuality, personal safety, and train Wi-Fi in the same part of the questionnaire was setting the research up for killing the Wi-Fi. Obviously people prefer not being stabbed on the train above Wi-Fi and "arriving at the destination on time" is the main purpose of the whole field.
"We want commute to be worse and more expensive" is a bizarre take to have as the department of transport.
Is private transport (i.e., the construction of roads) even financially sustainable? I did a quick search and it appears in the UK the amount brought in from vehicle tax each year is less than the amount spent on roads (https://www.gov.uk/government/statistical-data-sets/rde01-ro...).
Transport is massively profitable for the economy but not for the transport companies themselves, they simply can't capture enough of the value generated to turn a profit. If they try then the level of travel disproportionately goes down to make them lose even more money. It's why most/all of a public transport needs to publicly funded in some way and saving a penny or two per ticket on WiFi will cost far more to the economy than it saves.
There is one notable exception. Japan's train system is profitable depending on how you look at it. The trains themselves are a "loss leader" in that they actually lose money per train ticket, but the company that runs the trains also owns the terminals (lots of retail space) and the commercial real estate around them, which is where the money is made. They are in effect a commercial real-estate holding company that's doing transport on the side to beef up their portfolio.
Well, based on my experience dealing with in-train coverage both working at a telco and as a traveler, if their hardware was deployed nearly ten years ago, then it is quite likely that their passengers _are_ using their own data plans instead of it.
Manchester to London tomorrow morning is £184 in standard class, £255 in first class. People/businesses who can afford that are choosing it because it's better than driving.
(Both tickets are significantly cheaper if booked in advance.)
I any case, there's voice service over almost the entire country. It's the fast data coverage that can be a little patchy on some more remote bits of railway lines.
I have raised this issue before, let me rehash it. Railways is probably the worst mode of travel in the history of mankind. I have not found any examples of large scale, passenger railway that is a sustainable business.
But people continue to hype it up and it's always "Let's spend more billions of other people's stolen money into this, it will definitely work out then!". It is never "Let me invest billions of my own money into this railways system."
If something has never worked, it is unlikely to start working now, unless you change the economics fundamentally. I haven't seen that in any system. The wastage is immense, and the money is better spent on things that actually provide more value for less cost.
To conclude, they should remove the entire railways instead of removing wifi from the trains.
Nobody looks at the billions spent on roads and thinks "ooh these are unprofitable let's get rid of them", so why do people/governments constantly do that for rail?
The reasoning people use to justify the spending on roads is that the economic value people and goods transported on said roads outweighs the costs of building and maintaining them. Trains are more efficent than cars on this simply bu economics of scale: one long tube with metal wheels carrying 2000 people into the city moves a lot more than a freeway of the same area.
Then why are railways always unprofitable while car companies are worth billions in the free market?
> Nobody looks at the billions spent on roads and thinks "ooh these are unprofitable let's get rid of them", so why do people/governments constantly do that for rail?
Because roads are profitable and make sense. Rail does not.
Also, railways are at least 100x more expensive than roads. And the entire infrastructure has to be subsidised. With roadways, you just build the surface, vehicles are wholly privately owned. That means trillions less spent from the taxpayer's pocket.
The track, electricity supply, signalling system etc in Britain is owned and maintained (and extended) by a public company, although it contracts out a lot of the maintenance work.
I am curious what you think a fair price for driving a car on the motorway from, say, London to Leicester (exactly 100 miles) should cost. Currently it's free, although fuel is taxed, perhaps around £5 for that journey.
HN guidelines:
> Eschew flamebait. Avoid generic tangents. Omit internet tropes.
> Please don't pick the most provocative thing in an article or post to complain about in the thread.
Second, there are profitable passenger railways - a very famous example is Japan's high speed rail system.
No, the only reason Shinkansen exists is because of government subsidies, both direct and via interest-free "loans" without any repayment terms.
> First, there are other things to optimise for other than financial profitability
You don't have to optimise for financial profitability. If the product or service you provide is valuable to people, it will automatically bring in profits. If not, then you either go out of business, or subsist on continuous subsidies, e.g. every single passenger rail system on the planet.
Nope. Railways are very capital intensive, thus by definition require a long time to reimburse those initial costs. So if companies never invest in any further infrastructure, at some point they'll become profitable and remain so (see: all american cargo rail companies that haven't done anything but the bare minimum infrastructure wise for decades). But demand isn't static, therefore to actually serve the needs of the population the infrastructure needs to evolve - more routes, faster ones, newer and better trains, etc. which requires constant capital. Also, providing a valuable service doesn't mean that it can be directly financially profitable, and this is a perfect example - the indirect benefits for the economy, environment and people (because more people can access more places for work, and emissions can be drastically reduced, etc.) can and usually drastically outweigh the direct financial cost of running the route.
> No, the only reason Shinkansen exists is because of government subsidies, both direct and via interest-free "loans" without any repayment terms
Like literally any public infrastructure ever? That's literally one of the major points of a government, to build infrastructure everyone needs and will use. Roads, bridges, airports, railroads are capital intensive and rarely if ever profitable on their own in the first decades of their use. Unless you're advocating for personal teleportation devices for everyone, I'm not even sure what your argument is supposed to be; but then again, you said "stolen momey" when refering to taxes, which makes you unqualified to weigh in on any topic around infrastructure, urbanism, government spending or organisation, etc.; anyone with such a misunderstanding simply does not know enough to be a useful participant in a discussion.
So you're saying passenger rail can never provide enough value to people to justify the constant influx of capital? Cool, thanks for making my point.
> providing a valuable service doesn't mean that it can be directly financially profitable
No, it literally means exactly that.
> Like literally any public infrastructure ever?
You literally said that Japanese high speed rail is profitable. Why change tracks (pun unintended) when I point out the inaccuracy?
> That's literally one of the major points of a government, to build infrastructure everyone needs and will use.
No, but let's not diverge from the fact that rail is the worst, most inefficient mode of transport in history.
> Roads, bridges, airports, railroads are capital intensive and rarely if ever profitable on their own in the first decades of their use.
Yes, and out of those only railroads are the one which are never profitable, even after 10 decades of constant technological advancement.
> Unless you're advocating for personal teleportation devices for everyone
I'm not advocating for anything except not wasting taxpayer money on inefficient, unsustainable transportation.
> anyone with such a misunderstanding simply does not know enough to be a useful participant in a discussion.
Classic cope when you can't attack the argument, just resort to insults.
Can not, or more likely should not receive enough money from consumers to make for the constant influx of capital needed for it to remain as useful as possible. The main goal is to be as useful and practical as possible, not to make the most money possible - a fully business class with a spa and golden jacuzzi train might be profitable (like the Orient Express was as a luxury thing) but again, the goal of passenger rail is to be accessible and practical to most people. If your only way of measuring productivity is financial profit, yeah, I see you can't understand that.
> You literally said that Japanese high speed rail is profitable. Why change tracks (pun unintended) when I point out the inaccuracy?
Because you elaborated on your definition of profitability (can't have received any money at all from a government, ever), which no infrastructure project from the past few decades really meets. In my view the Japanese rail system is profitable after some initial well needed help; and again, if you apply your own (bad) measurement stick to any infrastructure, you'll find that nothing is really profitable. As for your argument that rail is the least profitable, I'm fairly sure if you do the math you'll see that Japanese rail has been less unprofitable (has needed less subsidies and bailouts) than Japanese airlines over the same time.
> Yes, and out of those only railroads are the one which are never profitable, even after 10 decades of constant technological advancement.
Airlines literally need bailouts on a what, decade schedule? But sure, railroads are "never profitable" because they get subsidies and need continued improvement to get better.
> I'm not advocating for anything except not wasting taxpayer money on inefficient, unsustainable transportation.
How is rail not sustainable? It's the cleanest and environment friendly method of transportation by far, which is very sustainable.
> Classic cope when you can't attack the argument, just resort to insults.
I am attacking your argument, but that doesn't mean I can't dismiss your entire position as one of ignorance (which, again, doesn't mean I'm not trying to explain to you why you're wrong). I cannot take libertarians seriously, only someone profoundly privileged or ignorant can have such views on the world.
Can't you say that for many "successful" car companies too? They're only around because of government bailouts and because of indirect government subsidies like road construction.
Sure, they will find a way to waste any money gained from this but the basic idea seems sound.
It should even be a win for the mobile providers too as it uses the spectrum more efficiently since less is wasted on interference.
Not really. In tunnels and remote areas (I mean middle of nowhere train tracks with fields, hills, cows as the only things within a multi-km radius) coverage can be extremely spotty. Even on very highly used high speed routes like Paris-Lyon or commuter routes like Paris - Melun there are blind spots (and big ones depending on the provider).
Public Wi-Fi "login" is absolutely awful. It relies on captive portals which 1) provide no security (the traffic is still unencrypted and you can take over someone's session by spoofing their MAC address, making the whole "auth" pointless), 2) don't actually fulfil any law-enforcement objective (malicious users will just provide fake details on the signup page anyway) and 3) break all kinds of applications and ultimately ruins the user experience to a stage where people just don't bother and tether to their mobile instead.
The solution is to either amend the law to make the whole "login" thing unnecessary (criminals will "login" with fake details anyway), make it open to everyone (so connecting immediately gives you unrestricted internet access with no captive portal, thus devices can auto-connect) and use some per-MAC-address and maybe per-domain traffic shaping to throttle down heavy users or heavy services (video streaming, etc).
Alternatively, if "login" or payment is needed, use WPA2/3-Enterprise (with EAP) which allows usernames and passwords (so the OS can handle the auth instead of a captive portal) with a single (potentially government- or telco-consortium-operated) login that works across all public networks so Wi-Fi actually works seamlessly and effortlessly. The Virgin Media Wi-Fi in the London Underground actually works similarly where users or any major carrier can seamlessly connect via EAP-SIM/EAP-AKA, never see a login page, and their device automatically roams on and off the network as needed.
In that case, adding WiFi inside the train, with the antenna outside the accidental Faraday cage, is worthwhile.
If it’s one less thing to distract the TOCs from you know, operating trains, then I say get rid of the Wifi.