I've always thought night trains are a fantastic, sustainable alternative to short-haul flights, but they're often held back by a lack of privacy, comfort, and poor economics due to low passenger capacity.
I became overly fascinated with this puzzle. I view it as a kind of night train Tetris (my wife less charitably calls it "sardinology"). I spent way too much time learning about and sketching various layouts, trying to figure out how to fit the maximum number of private cabins into a standard railcar, while making them attractive for both day and night travel.
This eventually led to a physical workshop (in Berlin) and a hands-on rapid prototyping process. We've built a series of full-scale mockups, starting with wood and cardboard and progressing to high-fidelity versions with 3D-printed and CNC-milled parts, with various functional elements.
Hundreds of people have come in to test our various iterations, because you can't test ergonomics or comfort by looking at renderings (although we did create a bunch of nice ones).
The link goes to our home page showing our approach and some of the thinking behind them. It’s been a lot of fun working on this puzzle, and we're excited to share what we've come up with. We hope you think it's cool too and would love to hear your thoughts.
Sleeper trains are held back by flying getting subsidised heavily by not having kerosene taxed, and national governments giving airports effectively unlimited room to grow; happily externalising the environmental cost. Why take a train if you can fly for a fraction of the cost?
Trains in general are held back by governments not investing in rail infrastructure, because the pork barrel of another motorway link is so hard to resist (and we're not properly maintaining these either).
Sleeper trains are held back, because cross-boundary collaboration between the various semi-national rail companies is tough (for Europe).
Sleeper trains are held back, because there is a lack of modern rolling stock. Not completely new concepts; just up-to-date sleeper wagons (the ÖBB has the leading edge here now with their new wagons).
There is room for improvement in the wagon designs, but it is almost irrelevant in the face of the other challenges.
You start off by essentially claiming the unit economics of night trains being too poor compared to aviation is the largest hurdle, then finish off by claiming that unit economics are not that major issue.
Our perspective is that with much improved unit economics, the problem overall becomes much more easily solvable. You can compete with aviation on price. You can pay for prioritized track access. You can operate trains privately without direct involvement of national operators.
Finally, the refurb approach skirts the rolling stock bottle neck.
Hi! What's your perspective on the shortage of manufacturing capacity for night train rolling stock? Last I heard ÖBB can't build them fast enough for demand, and few other companies are able to actually produce these? Are you planning to build your own industrial manufacturing capability for this? And what about 2nd level suppliers?
Yes, the bottle Beck for night trains is rolling stock.
The plan is to initially target refurbished standard passenger couches. We have a large research and development consortium to develop the technology and adaptation of coaches.
The refurb itself can be done by many companies, we don’t need a large supplier like Siemens.
As for the ÖBB nightjet, there’s talk about shifting the last ordered nightjet to rail jets (for day travel). Speculation is that the low capacity of the Trainsets result in poor unit economics.
I gotta say I love this? Are you in contact at all with the folks advocating for more night trains across the EU https://back-on-track.eu/ ? Seems like it could be pretty mission aligned.
Btw I run a weekly newsletter about urbanism and while your trains may not be exactly related I think it's cool enough that we'll feature it in the upcoming week! https://urbanismnow.substack.com/
Great work so far, but I note you are looking at a custom double-decker railcar with what looks like a very large loading gauge. My understanding is that for most of Europe, double deckers are not used due to loading gauge limits imposed by tunnels, bridges and so on. I presume your planned routes take that into account? Is that one of the reasons your mocked journey planner app doesn’t include the UK?
When traveling I often consider trains, and especially overnight trains. It's by far the most comfortable way to travel. Innovation in this space is a good thing.
While I'm aware that feature creepy is the enemy here, I would suggest a way to "combine " two pods for those traveling as a couple. If I'm traveling with my wife we don't want to be in "separate pods".
A retractable "wall" between 2 pods would be fine. It doesn't have to be elaborate, but you wanna point to something outside and say 'look at that' etc.
We considered connecting pods. The orientation „behind“ each other makes interaction difficult.
They way this issue was „resolved“ more less naturally during testing is that the pods all have The same orientation, so pods across the aisle approximately face each other. In our lab we had two iterations of the pods set up to face each other, and tester and testee interacted quite naturally —- once we set up our test rig like that, the questions „what about couples“ reduced a lot, most understood the vis-a-vis intuitively.
Our bigger cabins have two ppl versions, but a lot (if not most) travel is individua anyway, especially if night trains will be used for work travel.
I've long joked that modules should be made easy to swap with a forklift. Trains are usually full of small defects that aren't serious enough to take them out of service.
If they are comfortable you could rent out the cabins when not in use either fitted on the train or not. You could also retire the units there.
You could make a platform only and make it easy for others to design modules in a broad price range. Maybe most modules should be in storage until booked.
You could park the "hotel" module on the destination and put it back on the train for the return trip.
I sometimes her proposals like that, and it sounds kind of attractive - you get into your pod and forget everything until you arrive.
But in a sense, night trains are already like that. Since they can stop at multiple places, you can depart and arrive downtown. In the meantime you’re in your cabin and forget everything.
Entering the train „with“ the pod instead of just yourself is gonna make boarding and alighting take forever, and the logistics of storing and moving the pods are a nightmare. It’s going to reduce capacity by a lot because you cannot optimize the layout and every pod needs to be „insertable“ as a whole. (3x reduction in capacity means 3x increase in ticket cost).
Homologation is going to be a nightmare - in Europe, realistically, it’s gonna take more than 10 years or develop something like that. You need a new infrastructure because right now stations are for people - that’ll probably take 20 years (in Europe).
Besides, sleeper trains really are modular already, using existing standard rail infrastructure. Wagons are the natural unit of a train consist.
Ideally, we end up with a big enough network of sleeper lines that it makes sense for ÖBB, European Sleeper, or whichever operator to have the luxury of taking some wagons out for maintenance and standby, and even scale up and down depending on the season.
I love the concept and the renders, but I wonder... is the spatial optimization really what's holding train travel back? If you decrease passenger density in order to increase privacy and comfort, do you then have a corresponding increase in ticket prices?
In places with good train travel, it seems like they already have several cabin classes, from sardine seats (still luxurious compared to air travel) to private cabins (at several multipliers of price). Pod style rooms would presumably be cheaper than that, but still a lot more expensive than a seat?
Then in places without high speed passenger rail, like the US, this wouldn't really be able to address the major problems with train travel (slowness, lower priority than freight, low reliability, etc.).
Under what scenarios would using pods instead of cabins be more economically viable? And could these be retrofitted into existing sleeper cars, or would they have to build entirely new trains?
My experience on the Caledonian sleeper, in a 'room' was quite cramped in all senses (I am over 6 feet tall), and quite expensive aswell. If that had been optimized better, I would have enjoyed it more.
These renders do not make me feel as if I am even considered as a desirable passenger at 200 cm.
The new ÖBB wagons seem much more practical (and currently exist). A normal sleeper train wagon with stacked beds in compartments is fine for me. This origami concept looks claustrophobic, and the sleeping positions seem to allow for no room for the normal movements you make in your sleep, let alone getting out to take a piss or something.
Well, you cannot assume ergonomics from renderings. That’s why we run extensive testing. We tested on a large range of people. In the pods there’s a fall off in perceived comfort around the 95th percentile. Even then, the feedback is overall quite good.
In the larger pods, there’s actually an uptick in evaluation for taller people. Testers were often surprised how well it works.
All beds have at lest 2m, although there are different degrees of becoming smaller at the foot end — just like in aviation business class (with ticket prices 1.5 orders of magnitudes higher).
Clearly different markets have quite different requirements and comparisons to air travel.
For example a "night train" maxes out around 12 hours. A train from 6pm to 6am is functionally equivalent to a 8pm flight, arriving at midnight, checking into a hotel, getting some sleep etc.
How far you can go in that 12 hours (give or take) depends on the speed of the train etc. In Europe you can go to a lot of places in 12 hours. In the US not so much.
Much longer and other factors come into play. You have to balance the time cost of "getting there" to the time benefit of "being there".
But thats OK. This solution doesn't have to work everywhere. It can start where it works well and grows from there.
That is like comparing range in gas cars and EVs. Some do that but there are other major benefits.
The lengths I will go through to avoid air travel is much higher than a 1:1 ratio in comparable time. When I have to get the cattle treatment I prefer cattle cars over cattle cans.
And even with 1:1 remember that layovers are a completely different beast. If Münich was a hub between Northern an southern Europe I would be happy to spend a well rested day before continuing on. Especially in spargel season!
...but only a fool does not fear German railroads. They could really learn from the Austrians.
The reason night trains are not a thing is because there is no real network. Looking for tickets in Europe it is often once or twice a week departures on specific routes. No real good north south interconnected corridor from Scandinavia.
And as a proper geek I have even sought them out but often found them sold out.
You yourself mention the trade off — the „sardine class“ is a six-people per cabin couchette (60 ppl per railcar). The private class is luxury (10-17 cabins per railcar).
We got 65 private pods or close to 40 little single cabins - in a refurbished railcar. In a new car it would be more.
We put together some explanation of the economics and the difference between old and new cars: https://luna-rail.com/approach
What's up with the super high carbon intensity estimate of your trains? Is it because they're running at night and can't use solar? Does it include embedded carbon in the train? Or is Germany's grid just that dirty?
15g/km is very low. The number is dominated by energy mix and infrastructure. Deutsche Bahn publishes lower number, but they only look at operation (not infrastructure), and assume „green energy“ rather than the energy mix of the country.
Hi HN, I'm Anton, founder of Luna Rail.
I've always thought night trains are a fantastic, sustainable alternative to short-haul flights, but they're often held back by a lack of privacy, comfort, and poor economics due to low passenger capacity.
I became overly fascinated with this puzzle. I view it as a kind of night train Tetris (my wife less charitably calls it "sardinology"). I spent way too much time learning about and sketching various layouts, trying to figure out how to fit the maximum number of private cabins into a standard railcar, while making them attractive for both day and night travel.
This eventually led to a physical workshop (in Berlin) and a hands-on rapid prototyping process. We've built a series of full-scale mockups, starting with wood and cardboard and progressing to high-fidelity versions with 3D-printed and CNC-milled parts, with various functional elements.
Hundreds of people have come in to test our various iterations, because you can't test ergonomics or comfort by looking at renderings (although we did create a bunch of nice ones).
The link goes to our home page showing our approach and some of the thinking behind them. It’s been a lot of fun working on this puzzle, and we're excited to share what we've come up with. We hope you think it's cool too and would love to hear your thoughts.
Sleeper trains are held back by flying getting subsidised heavily by not having kerosene taxed, and national governments giving airports effectively unlimited room to grow; happily externalising the environmental cost. Why take a train if you can fly for a fraction of the cost?
Trains in general are held back by governments not investing in rail infrastructure, because the pork barrel of another motorway link is so hard to resist (and we're not properly maintaining these either).
Sleeper trains are held back, because cross-boundary collaboration between the various semi-national rail companies is tough (for Europe).
Sleeper trains are held back, because there is a lack of modern rolling stock. Not completely new concepts; just up-to-date sleeper wagons (the ÖBB has the leading edge here now with their new wagons).
There is room for improvement in the wagon designs, but it is almost irrelevant in the face of the other challenges.
You start off by essentially claiming the unit economics of night trains being too poor compared to aviation is the largest hurdle, then finish off by claiming that unit economics are not that major issue.
Our perspective is that with much improved unit economics, the problem overall becomes much more easily solvable. You can compete with aviation on price. You can pay for prioritized track access. You can operate trains privately without direct involvement of national operators.
Finally, the refurb approach skirts the rolling stock bottle neck.
Hi! What's your perspective on the shortage of manufacturing capacity for night train rolling stock? Last I heard ÖBB can't build them fast enough for demand, and few other companies are able to actually produce these? Are you planning to build your own industrial manufacturing capability for this? And what about 2nd level suppliers?
Yes, the bottle Beck for night trains is rolling stock.
The plan is to initially target refurbished standard passenger couches. We have a large research and development consortium to develop the technology and adaptation of coaches.
The refurb itself can be done by many companies, we don’t need a large supplier like Siemens.
As for the ÖBB nightjet, there’s talk about shifting the last ordered nightjet to rail jets (for day travel). Speculation is that the low capacity of the Trainsets result in poor unit economics.
I gotta say I love this? Are you in contact at all with the folks advocating for more night trains across the EU https://back-on-track.eu/ ? Seems like it could be pretty mission aligned.
Btw I run a weekly newsletter about urbanism and while your trains may not be exactly related I think it's cool enough that we'll feature it in the upcoming week! https://urbanismnow.substack.com/
Yes, we’re connected with back on track! I’m a day-one member of the German branch. A bunch of their members came and participated in our tests.
We’re participating in their night train conference in September in Berlin, probably bringing a physical prototype.
Really cool concept. I wish we could have something like this in the US one day.
Wanted to report a small typo, In the 3D model index menu, "Uppder" can be found. I assume this was supposed to read "Upper", as in "above".
I hope to one day ride such a system when I visit europe, best of luck with your project.
Great work so far, but I note you are looking at a custom double-decker railcar with what looks like a very large loading gauge. My understanding is that for most of Europe, double deckers are not used due to loading gauge limits imposed by tunnels, bridges and so on. I presume your planned routes take that into account? Is that one of the reasons your mocked journey planner app doesn’t include the UK?
We have them in Switzerland. Often had to change train at the border and never thought it could be a technical limitation.
The „current“ concept is for refurbished standard passenger coaches, which is compatible everywhere in Europe except UK.
The „long term“ concept is built around to be compatible with profiles UIC GB and G2, which work in most of Europe.
When traveling I often consider trains, and especially overnight trains. It's by far the most comfortable way to travel. Innovation in this space is a good thing.
While I'm aware that feature creepy is the enemy here, I would suggest a way to "combine " two pods for those traveling as a couple. If I'm traveling with my wife we don't want to be in "separate pods".
A retractable "wall" between 2 pods would be fine. It doesn't have to be elaborate, but you wanna point to something outside and say 'look at that' etc.
We considered connecting pods. The orientation „behind“ each other makes interaction difficult.
They way this issue was „resolved“ more less naturally during testing is that the pods all have The same orientation, so pods across the aisle approximately face each other. In our lab we had two iterations of the pods set up to face each other, and tester and testee interacted quite naturally —- once we set up our test rig like that, the questions „what about couples“ reduced a lot, most understood the vis-a-vis intuitively.
Our bigger cabins have two ppl versions, but a lot (if not most) travel is individua anyway, especially if night trains will be used for work travel.
Cool idea! Small typo on the landing page, in the media section it says "german nationa media" without the l
What do you see as the benefits over Nightjet’s Mini Cabins?
Also, interesting to see just after the launch of https://noxmobility.com/ which is targeting the same market.
ÖBB Mini cabins
Luna Rail NoxI've long joked that modules should be made easy to swap with a forklift. Trains are usually full of small defects that aren't serious enough to take them out of service.
If they are comfortable you could rent out the cabins when not in use either fitted on the train or not. You could also retire the units there.
You could make a platform only and make it easy for others to design modules in a broad price range. Maybe most modules should be in storage until booked.
You could park the "hotel" module on the destination and put it back on the train for the return trip.
I sometimes her proposals like that, and it sounds kind of attractive - you get into your pod and forget everything until you arrive.
But in a sense, night trains are already like that. Since they can stop at multiple places, you can depart and arrive downtown. In the meantime you’re in your cabin and forget everything.
Entering the train „with“ the pod instead of just yourself is gonna make boarding and alighting take forever, and the logistics of storing and moving the pods are a nightmare. It’s going to reduce capacity by a lot because you cannot optimize the layout and every pod needs to be „insertable“ as a whole. (3x reduction in capacity means 3x increase in ticket cost).
Homologation is going to be a nightmare - in Europe, realistically, it’s gonna take more than 10 years or develop something like that. You need a new infrastructure because right now stations are for people - that’ll probably take 20 years (in Europe).
Besides, sleeper trains really are modular already, using existing standard rail infrastructure. Wagons are the natural unit of a train consist.
Ideally, we end up with a big enough network of sleeper lines that it makes sense for ÖBB, European Sleeper, or whichever operator to have the luxury of taking some wagons out for maintenance and standby, and even scale up and down depending on the season.
I love the concept and the renders, but I wonder... is the spatial optimization really what's holding train travel back? If you decrease passenger density in order to increase privacy and comfort, do you then have a corresponding increase in ticket prices?
In places with good train travel, it seems like they already have several cabin classes, from sardine seats (still luxurious compared to air travel) to private cabins (at several multipliers of price). Pod style rooms would presumably be cheaper than that, but still a lot more expensive than a seat?
Then in places without high speed passenger rail, like the US, this wouldn't really be able to address the major problems with train travel (slowness, lower priority than freight, low reliability, etc.).
Under what scenarios would using pods instead of cabins be more economically viable? And could these be retrofitted into existing sleeper cars, or would they have to build entirely new trains?
My experience on the Caledonian sleeper, in a 'room' was quite cramped in all senses (I am over 6 feet tall), and quite expensive aswell. If that had been optimized better, I would have enjoyed it more.
These renders do not make me feel as if I am even considered as a desirable passenger at 200 cm.
The new ÖBB wagons seem much more practical (and currently exist). A normal sleeper train wagon with stacked beds in compartments is fine for me. This origami concept looks claustrophobic, and the sleeping positions seem to allow for no room for the normal movements you make in your sleep, let alone getting out to take a piss or something.
Well, you cannot assume ergonomics from renderings. That’s why we run extensive testing. We tested on a large range of people. In the pods there’s a fall off in perceived comfort around the 95th percentile. Even then, the feedback is overall quite good.
In the larger pods, there’s actually an uptick in evaluation for taller people. Testers were often surprised how well it works.
All beds have at lest 2m, although there are different degrees of becoming smaller at the foot end — just like in aviation business class (with ticket prices 1.5 orders of magnitudes higher).
Clearly different markets have quite different requirements and comparisons to air travel.
For example a "night train" maxes out around 12 hours. A train from 6pm to 6am is functionally equivalent to a 8pm flight, arriving at midnight, checking into a hotel, getting some sleep etc.
How far you can go in that 12 hours (give or take) depends on the speed of the train etc. In Europe you can go to a lot of places in 12 hours. In the US not so much.
Much longer and other factors come into play. You have to balance the time cost of "getting there" to the time benefit of "being there".
But thats OK. This solution doesn't have to work everywhere. It can start where it works well and grows from there.
That is like comparing range in gas cars and EVs. Some do that but there are other major benefits.
The lengths I will go through to avoid air travel is much higher than a 1:1 ratio in comparable time. When I have to get the cattle treatment I prefer cattle cars over cattle cans.
And even with 1:1 remember that layovers are a completely different beast. If Münich was a hub between Northern an southern Europe I would be happy to spend a well rested day before continuing on. Especially in spargel season!
...but only a fool does not fear German railroads. They could really learn from the Austrians.
The reason night trains are not a thing is because there is no real network. Looking for tickets in Europe it is often once or twice a week departures on specific routes. No real good north south interconnected corridor from Scandinavia.
And as a proper geek I have even sought them out but often found them sold out.
They cost optimized themselves to obliteration.
You yourself mention the trade off — the „sardine class“ is a six-people per cabin couchette (60 ppl per railcar). The private class is luxury (10-17 cabins per railcar).
We got 65 private pods or close to 40 little single cabins - in a refurbished railcar. In a new car it would be more.
We put together some explanation of the economics and the difference between old and new cars: https://luna-rail.com/approach
What's up with the super high carbon intensity estimate of your trains? Is it because they're running at night and can't use solar? Does it include embedded carbon in the train? Or is Germany's grid just that dirty?
Where do you see the carbon intensity? I just see a plane above a train with a single "250g" figure.
I see 15g CO2(eq)/km in the bottom left-hand corner
15g/km is very low. The number is dominated by energy mix and infrastructure. Deutsche Bahn publishes lower number, but they only look at operation (not infrastructure), and assume „green energy“ rather than the energy mix of the country.
The 15g is a good estimate, see discussion here: https://back-on-track.eu/de/klimawirkung-von-nachtzuegen-neu...