Trains are already crazy efficent both fuel and operational people wise - in fact this negates one of railroading great efficencies, slack (that 2 mile freight train already starts one car at a time). Also subtracting, cars are going to need to bunch up anyway at passing sidings to let cars go by opposite direction. Besides, the real cost is railroading is MOE and MOW (including land taxes here) and this solves neither of those issues. Could be interesting for locals and LCL traffic but those haven't been big parts of railroading in decades.
Rail is insanely efficient over long distances but these seem to be optimized for shorter route with a 500 mile limit ignoring battery degradation. For rural lines that might see 10 trains a day increasing utilization via ultra short trains on local routes could be very profitable for a railroad as long is it doesn’t require people or interfere with existing traffic.
Basically if a tiny spur for loading and unloading doesn’t need to fit a full sized train they you can cheaply add a lot of stations.
I think the true value of this isn't replacing trains during the long haul portion, but rather the "last mile". The logistical infrastructure for getting freight from train depot to other depots is about as efficient as it can get BUT getting that freight to its final destination is still inefficient. A 2 mile long train needs to tediously be loaded onto trucks, which will all burn gasoline to get to where-ever they need to go.
In Switzerland a lot of factories and warehouses have train lines built into the infrastructure to try to solve this issue, but its not always feasible to a) build that infrastructure out, especially in the US, and b) use a train to transport a few containers to its destination.
If these electric train cars could be built to interface with normal train cars, they could still take advantage of the biggest efficiency of freight trains, while also helping solve the last mile problem by splitting off from the train to get closer to their final destination.
In large cities you could lay more tracks specifically for these cars to get closer to major industry areas. Land rights are an issue but you could lay track in roads. Chicago has some remnants of street running trains, but the biggest issue is a train that's ~200 yards long is a lot more disruptive to traffic than something that's the size of a semi truck.
That's what locals are, the last mile (or hundred mile) trains that get made up out of cars that have moved long distances already. Railroads hate thise trains. The money is in unit trains, all one commodity going from one place to one other place over a very long distance - the exact opposite profile of locals. Railroads started making this transition in the 60s and 70s as they shed branches and secondary mains and excellerated into the 80s and 90s by spining off shortlines that would focus on local traffic.
I can think of numerous political/social/business reasons for this to go wrong, but can't really think of any way in which this isn't more efficient than the status quo and so seems inevitable on some timescale.
Before, long train that has to move as one unit. After, long train that can optionally split and recombine at will with other long trains coming and going from different places.
You've essentially described the trucking industry, without the need to figure out linking or being limited to rail routes.
Without the train length, you lose the efficiency. The sum of added traction drivers and energy storage on each car exceeds any savings you might gain by autonomous operations or possibly getting rid of the traditional locomotives.
Convoy based self-driving tech on interstates is roughly mature enough as-is to operate as a drafting, self assembled train, if there was a desire/demand to implement it.
But that goes back to political/social/business reasons.
one of railroading great efficiencies, slack (that 2 mile freight train already starts one car at a time)
Is that right? Maybe before roller bearings, when overcoming starting friction was a huge problem.[1] But all US railroad cars in interchange have had roller bearings since about 1992.
Normal starting procedure is to move the engine very slowly until the end of train device reports motion, indicating all the slack has been taken out. Then apply power.
Exactly,slack is the space give to each car if the train to move before the next car has to. Starting cars rolling is the most energy expenssive thing the motive power has to do. So once you take the slack out on a single car it's pretty cheap to keep it rolling (Issac Newton to the rescue!) - then on to the next car. So a train starts rolling one car at a time as the slack gets taken out. Crazy efficent.
> Starting cars rolling is the most energy expenssive thing the motive power has to do.
Is it, really? It might require an not unconsiderable amount of force, but due to P = F x v, at low speeds even high forces don't require that much power. If anything, it's rather that you're friction-limited at that moment.
> Also subtracting, cars are going to need to bunch up anyway at passing sidings to let cars go by opposite direction.
Sounds like the cars are already “bunched up”:
> “We think our platoon sizes are ideally between ten and 50 cars,” Matt Soule, CEO of Parallel Systems, told Ars.
Edit: Also, you say:
> … [T]he real cost is railroading is MOE and MOW (including land taxes here) and this solves neither of those issues.
But the article mentions that the maximum utilization of a rail line is currently limited by how close trains can follow one another safely. If new technology can increase that limit, the capital costs of the rail line can be distributed across more units of freight.
But the article mentions that the maximum utilization of a rail line is currently limited by how close trains can follow one another safely.
Back in the 90's, Wisconsin Central was a major freight railroad in the midwest (I don't know if it still is), and was trying really hard to be allowed to automate most of its freight lines.
Part of the problem was that for legal reasons, it had to convince every little town along the way to buy into the idea of automated freight trains rolling through their towns. I went to a bunch of public hearings about it.
At the time, WC had already reduced its trains to one or two humans (I think just one), and it was really easy for local politicians to envision a voter's car being struck by a train at a crossing and dragged for 500 miles, since there was no engineer to see it happen.
WC's response was that it was already running automated freight trains in New Zealand with no problems because they had cameras on board.
It then mentioned that with automated trains, it could run more trains closer together. If you've ever lived near a railroad, whether small town or big city, stalled freight trains blocking grade crossings are a major problem. So the little towns didn't like hearing that automation meant more trains.
I don't know what the status is today, but by the time I left the midwest, WC had instead shelled out the money to upgrade the tracks and grade crossings through the little towns so that it could plow trains through without slowing down, achieving the efficiency it was looking for, without resorting to automation.
Wisconsin Central was my local railroad growing up, and they've since been bought out by Canadian National (in 2001). When I look at documentary footage from the 90s or before of trains coming through my town it's obvious how much effort they put into upgrading the tracks and removing as many grade crossings as they could. Driving a few extra blocks to get to a crossing was annoying when I lived on the other side of the tracks, but I can definitely appreciate the improved safety and increased speed of the trains compared to when Wisconsin Central originally took over the rickety old Chicago Nortwestern (/Fox River Valley RR) and Soo trackage. CN can speed trains through town at 35MPH so it's not usually too long of a wait.
How does this negate the efficiency from slack? Each bogie is sized to move one car and all the cars move independently. Isn’t that effectively slack? Unlike regular freight trains, it wouldn’t have the momentum of all cars in front helping to start any given car, but on the other hand, that momentum was still produced by the locomotive earlier in the starting of the train.
is MOE maintenance of equipment and MOW maintenance of way?
I thought the big inefficiency of rail was not physics, but capitalism and competition.
the cost of starting a trucking company is negligible and a truck can immediately use the shared resource roadway.
Meanwhile use of a railway system is governed by a central authority and shipping is not subject to competition in the same way. Additionally long segments of single-rail prevent efficient use of routes
Would be interesting if there was a two-way (two or more tracks) railway system where trains could hop on and hop off on a first-come first-served basis...
There are a lot of problems with this proposal, and it's also not clear what problem it's solving.
Battery life prevents using this for long hauls, but the idea of using these small pods that can easily split up on small industrial spurs doesn't make sense either - most of those non-main-line routes have manually thrown switches. So beyond the much higher unit economics of requiring 2 complex autonomous vehicles (and associated maintenance) per container instead of a basic hunk of steel with wheels, you still need to have a person travel with it to throw switches, or else upgrade every switch on every tiny, poorly maintained spur to be electrically operated.
There's also some strange claims by this company in the article, like the ability for these to carry double-stacks. If you look closely at current intermodal rolling stock, you'll notice that the containers are between bogies rather than above them. A double-stack container would sit too high to fit through standard tunnels and bridges if the containers were on top of the wheels. There's no possible way to accomplish this concept (a pod per end of a container) if you need the container to sit lower than the top of the wheels...
Maybe there's some potential with this idea, but I'm really not buying it based on what they're saying and showing so far.
To me this seems like a possible solution for the "last mile" of rail. My imported shipments from overseas end up getting delivered to a city that's 350 km away just because there are no large scale rail depots any closer. Sadly, this may be too late as most of the smaller rail lines nearby have been ripped out of the ground. The decimation of rail infrastructure compared to what existed 100 years ago has already happened, so they have quite an extreme uphill battle to gain any acceptance.
But then you need a transfer facility with this big gantry crane at the boundary of every “last mile” to get the container off the traditional train and onto the autonomous team.
A "big crane" is nothing compared to, as the article explains, 3 mile long trains. In fact it's also nothing compared to the space needed for normal temporary container storage.
The benefit of rail is that you exchange the enormous capex of shaping and planting metal rails for only having to run and maintain a single locomotive that pulls simple boxes with wheels that can move staggering amounts of tonnage efficiently.
This system requires both building more infrastructure to leverage the granularity of the cars, more infrastructure to keep a massive fleet of batteries charged, and more maintenance as each freight car now has a drive system.
If this already existed someone would pitch leveraging the economies of scale where you'd build a huge car that pulls all the other ones and you only need to keep 1 battery charged... which is what we have right now...
Actually it turns out rail lines have fairly low capex compared to many other transport types. It makes sense once you think about it, but I was surprised too.
Unfortunately, for historical reasons the US railroads are private and pay many sorts of taxes, and the roading industrial complex is largely publicly funded.
Many lines used to have double or quadruple or even electrification but are now single unpowered tracks, because the additional improvements counted towards property tax assessments. It's just not a level playing field.
I wasn't sure. Intuitively rail would seem to be cheaper, it looks like it should be a lot easier to lay a rail line than build a road. But rail lines need a lot more fill or bridging over low areas since heavy freight trains cannot climb grades of more than a few percent. On the other hand, rail switches and diamonds have got to be way less expensive than highway interchanges.
At least for passenger lines the single track of heavy rail has a >30x higher route capacity than a single freeway lane (https://en.m.wikipedia.org/wiki/Route_capacity). So a fair comparison would be a single line of heavy rail to a 30+ lane freeway (one direction), making the rail about 1/10th the cost.
Assuming you’re referencing the graphic on that page, note that it’s discussing throughput of road vs rail in the city. Freeway is a very different kettle of fish.
That makes it cheaper if you can use it: a railway track has 20 times the capacity of a freeway lane.
(This is just the result of a basic Google search. It also says maintenance costs are "much less" for the railway.)
There are probably economy-of-scale issue too (we build a lot more roads), and different expectations of road vs. rail safety (a rail accident with a few injuries has national media coverage for days, upgrades costing millions may be rolled out to prevent a repeat; every day several people die on roads, the infrastructure is rarely improved).
This does not scale economically when regular rail already has a difficult time being profitable.
Switzerland has the most dense rail network in the world and is 100% electrified. However even before the pandemic SBB cargo was in the red. To reach our climate goals we need to increase the amount of cargo on rail (currently around 46% of cargo is by rail). There is no way around it but they will need government funding to reach those goals.
Such gimmicks have extremely high costs. Each on of those needs maintenance. A regular cargo car is quite dumb and can take a lot of abuse.
High speed, or even decently fast rail, needs track that is generally less curvy, maintained to a higher standard, etc. and that costs money. European cargo networks generally don't make money because track access fees are really high, particularly, if, say, you want to use that new 250kph tunnel through the Alps that cost billions of dollars.
American freight rails are profitable, but that's because they totally shit on passenger rail. The tracks are congested, and in extremely poor shape. But that's fantastic for freight rail; a lot of what moves by freight rail doesn't necessarily need to be truly quick.
China split the difference by building a whole new high-speed passenger rail network to free up space for freight trains on the older network. This seems to work out, but China Railways has a balance sheet that would make Westerners blanche.
The balance sheet would only make westerners miss a beat if they aren't aware of the balance sheet of the alternative, roads and cars, which is an order magnitude worse pound for pound.
No, there's not really a universe in which the high speed lines to Urumqi or Lhasa really make any (financial) sense. They literally do not sell enough tickets to pay to energize the wires.
Again, you're still doing what I said you shouldn't do. Compare those to highways to nowhere in the United States, or to many northern roafs in Canada. They are not used nearly enough to justify paying for their maintenance either, but we understand their necessity so that everyone is connected.
I believe Switzerland has nationalized their railroad network, much like the US has nationalized our road network. Its supposed to take a loss so that the rest of the economy benefits.
I wish I hadn't spent my time watching that. Argument through incredulity and mocking might be entertaining for some, but it's very low signal and all noise. Leaving this comment as a warning for others.
I watched only the second video and also find it poor. It basically just says that using this system to transport a train-load of cargo between two train terminals would use more energy than a train. Well, you don't say.
Also, as other videos on the channel it is arrogant and mocking.
I haven't watched the video, but based off of the title "Pod train" the argument is probably along the lines of this:
There is a futuristic theme in logistics that envisions humans or cargo to be shipped in "pods". This is in contrast to the more efficient way of shipping humans or cargo in larger vehicles like trains or busses.
These pod-ideas inevitable run into some bottle-necks or have to make trade-offs that in the end give them so much lower efficiency than the "traditional" way which is already the (almost) optimal solution to a given problem.
Examples of this are various incarnations of hyperloop, that do not actually solve all real-world problems existing solutions have to deal with, but rather only a subset of them - but they are better at this subset plus they look fancy in their renderings.
This makes it so these impractical solutions get a lot of hype, and money is spent on prototypes, but there is a lot of hot air and little substance.
TL;DR: Someone who actually knows trains points out that the designers eliminated all the benefits of rail while designing something that requires the entire existing rail network to be changed to support it (which will never happen).
I think the biggest nail in the coffin is that they are targeting short haul lines, which are often the least maintained and require human supervision to operate on for safety and logistical reasons. For example, short haul lines use manual switches - so every time two tracks split or come together you have to have a human present.
In the first video the author makes a lot of assumption around operating these things in the existing railways without changing how they are managed. It does not consider the cost savings and increased flexibility in having cars arrange themselves and route autonomously to their destinations.
Industrial warehousing/logistic robots are becoming very widespread and significantly changes the workflow and reduces human intervention.
You can't make a fair comparison by treating autonomous robots as if they were regular train cars.
The second video is simply mocking the boisterous style and claims of the ads.
I think they don't consider that because it's a rail system. Re-arranging cars on route would require many junctions, and would probably slow down the whole system when they separate. It's also a different problem that they are solving with warehouse robots - goods need to be transported individually to destinations. Trains solve mass transport. And besides - autonomous rail robots would still be bound to a rail, and thus can't really plan any fancy avoidance or path finding that a warehouse robot could.
> In the first video the author makes a lot of assumption around operating these things in the existing railways without changing how they are managed. It does not consider the cost savings and increased flexibility in having cars arrange themselves and route autonomously to their destinations.
Are they selling the bogie cars, or are they selling new rail infrastructure? If it's just the former then they have to work with the existing infrastructure, including manually switched tracks
> a lot of assumption around operating these things in the existing railways without changing how they are managed. It does not consider the cost savings and increased flexibility in having cars arrange themselves and route autonomously to their destinations.
Assumptions that are deeply rooted in reality. Rail is extremely cheap because it offers a fixed set of interconnection points with other modes of transportation and has extremely rigid standards that ensure at every step nobody has to stop and think about how to handle something.
This is like Tesla announcing the Cybertruck will be 25 feet wide and insisting that all roads nationwide be updated so they will work.
Bringing the ideas of packet based networking to the world of train based container freight seems interesting.
Question is will the gain in time and cost efficiency (by having each container move individually more or less fully automated) be bigger then the loss in energy and cost efficiency (each container needing it's own motor and energy storage sounds to be more expensive then one big locomotive)?
Packet based networks are great where it is possible to route around bottlenecks.
The rail network does not have that property.
Unfortunately, physics does not allow for packet collisions.
---
In terms of practical efficiency of motors: grades and loads and wind vary.
A self-powered rail care would need enough oomph to handle maximum load up a steep grade in a front quartering wind...most of that capacity would lie idle most of the time.
Packet based networks are not a good analogy because there is typically only one railway line the discrete container units can travel along - the magic of the internet is the multiple routes data can travel along concurrently to destinations.
When reading about this proposal the only thing that can justify it is scamming investors out of their money.
First of all this internet-packet-routing-like way of transporting goods would work only if almost every town in America had rail passing through it, or at least very near it, which isn't the case. Secondly if such a rail network were to be constructed it would pay off to just have it electrified and use these carts without a battery. Until that day what rail needs is maintenance, expansion and usage, not these silicon valley pipe dreams.
A lot of towns used to have rail service with spur lines or short lines. A lot of those have been turned into bike or walking trails. Good luck ever getting those back to their original use; re-provisioning rail service to these places will mean acquiring and clearing new routes.
I'm not sure if this is true everywhere, but I lived in a town with a "rail trail" and my understanding was that the railroad company was renting the land to the town, and could put it back into use whenever they wanted.
My very rural town has a rail line pass through, and it was once the main means of moving people and goods to neighboring towns. But those services were discontinued decades ago, and now the line only hauls lumber and similar bulk industrial goods.
I would love to see passenger and freight service available here again, and I feel that this could be something capable of providing that kind of service. Assuming the existing line owner would allow it. I doubt that happens without a truly epic business plan, which I’m not seeing here.
I want to be enthusiastic about this idea, but this will be an uphill battle on multiple fronts (as other commenters here have opined more pointedly).
It’s hard to hope for something that seems impossible. I think it a safer bet to expect horses genetically modified with narwhal DNA that produces a tusk on their foreheads.
I think my biggest concern is "platooning" instead of coupling. It only takes one car going slightly out of the sync to become a Very Big Problem (a lot of train accidents have happened via a full or part train going runaway).
It almost certainly would not get regulatory approval, particularly in the US where the FRA is very "my way or the highway" when it comes to new things, or standards that were NIH.
Autonomous vehicles carrying containers: [1] (Video is sped up). Most of the hardware for this already exists, and is in use at the more advanced ports for moving containers around within the port and sorting them. Major ports today are very automated.
The latest fad is "dry ports". This is a bigger version of what used to be called an "intermodal facility", a place where containers are transferred from trains to trucks. The long distance portion of the trip is by rail, and then there's a transfer to trucks for the last ten or hundred miles. That, in practice, is how this problem is solved.
I'm wondering whether it might become even more cost-effective if it can run off of third-rail electrification (the form factor makes overhead catenary impractical). That said, the issue about last-mile rails having been removed is very much a big issue. There was once a rail spur in downtown Chicago along Kinzie Avenue to serve printing plants at the Sun-Times and Tribune which is no longer maintained, although I'm not sure if there's a good use case for it even with these sorts of cars (IIRC it originally ran to loading docks off the river in an area that's now luxury condos). I remember a few years back going for a run nearby and watching workers tear up the rails from another rail spur off the Illinois Central¹ tracks.
⸻⸻⸻
1. Like many a Chicagoan, I have a tendency to refer to things by long-superseded names that no longer apply. I honestly could not tell you the name of the current owner of that right of way.
Third rail cannot deliver nearly as much power as overhead for voltage reasons. It might work for short lines where you don't need speed, but not for mainline
It mostly-works and is used in South East England (including South London + some bits), both for passenger and freight trains.
The top speed is about 160km/h. I don't know if the maximum weight or length of a freight train is reduced vs. overhead catenary.
However, it's more expensive to maintain and more dangerous to staff and the public. There's a very, very long term policy aim to convert everything to overhead catenary.
It might sort of work, but it's still less than ideal:
- A Class 92 in DC mode is limited to 4 MW, whereas on AC it can output 5 MW (and other, more modern four-axle electric locos manage 6.4 MW), and even that 4 MW capability already required significant infrastructure upgrades (I think they basically had to double the number of substations along the designated freight corridors between the Channel Tunnel and London)
- Anecdotally, even with just passenger trains on some heavily trafficked sections of route they're running into the problem that the current draw becomes so high that it becomes hard to distinguish between regular operations and an earth fault (and building even more substations isn't cheap, and if you compare it with the amount of substations you'd require for high voltage AC…)
- At switches and crossings you get unavoidable gaps in the third rail, and especially at the complex track layouts around some of the larger stations those can occasionally combine into gaps of substantial length. Multiple units with distributed traction (which have dominated passenger traffic on the Southern Region for a long time) can handle those better because there are pickup shoes distributed along the whole length of the unit, plus possibly a traction power bus connecting those pickup shoes together (at least within a single three-/four-/five-car unit). So all in all it's relatively unlikely that a multiple unit will find itself completely off the traction supply ("gapped").
With a freight train powered by a single locomotive on the hand, power supply to the locomotive will always be interrupted by anything but the very smallest gap in the third rail, which means that the tractive effort gets interrupted on each of those occasions (and to avoid jerking the couplings you want to turn off and re-apply power smoothly, and especially the former will make driving more complicated), you need to power the loco's auxiliary systems through all those interruptions, and if the train comes to an unscheduled stand within a junction (whether due to a technical fault, external influences like trespassers, a driver error, or whatever…) there is a much higher risk that the train ends up gapped and will then cause quite a bit of delay until you can somehow get it going again (if you're very lucky the track gradient might get the train rolling again, otherwise you either need to connect up some extensions leads up to the loco or else give it a push with another train/loco, all of which takes quite some time to set up).
It works just fine for the “L” here in Chicago. I'm not sure how the convoys would work in terms of power drain, my EE knowledge is pretty much zip, but since this would be for last-mile transportation by and large, I think it would be ok. The biggest concern would be public safety, I imagine.
For lightweight passenger cars it is okay. Freight is a lot more mass to get moving. Is also means you have a lot more expensive infrastructure costs as the relatively low voltages means more substations
Connect those things in line by hundreds to simplify management of the shipment and save energy. Put in front one big engine that put the whole thing in motion. Let the physics do the rest.
You just reinvented the container train.
Meanwhile in silicon valley, hey let's reinvent the less efficient and most complex way of doing that.
Also applies to Elon Musk Boring Company that is the dumbest, most dangerous, and less-efficient way of building a subway.
Actually, yes, you do save energy by connecting them. You reduce wind friction by having them draft each other, and you reduce weight by having one large motor instead of multiple smaller ones.
The wind friction is lowered in high-speed trains by making a flush body without interruption - and not by the connection itself. But this is besides the point.
Each car has its engines and brakes, there is no significant push/pull force on the connections and there is no bigger engine somewhere.
In the context of the current conversation the autonomous cars are electric and each car has its engines.
I'm not talking about high speed trains. I'm talking about medium speed cargo trains. Despite the lack of a flush body, each container does draft the following one, greatly decreasing drag.
Each car in a normal cargo train doesn't have its own engine. They are pulled by locomotives. This is more efficient than putting an engine in every car.
Passenger trains typically do have engines in every car, but we're not talking about passenger train here.
The tech is neat but the reasons this isn't already in use has to do a lot more with railroading company culture (operational familiarity, risk of losing business to a competitor) than with any particular shortcomings of the technology.
Right now, no one has to chaperone individual railcars (or bogies!), because trains of many railcars travel as a unit. This also makes track control / impact avoidance easier, regardless of the level of train control deployed on the track.
This may see more use in the EU, where EU-wide regulations are mandating all member states to separate ownership of their rail network from ownership of rail operators. Then, an adventurous operator may decide to trial this technology. But nonetheless, this is fairly unlikely, as rail slots are essentially priced by time occupied for the block, so it makes more sense to pack a train's worth of cargo into the reservation you paid for.
Even in the EU, you will want an engine in front and you'll want the train to be coulpled together. For one, the engine is the place where the train understands the gazillion signaling systems in the EU, and it's where the overhead power is converted to on train power. Both items cost like millions, it's not something you want to distribute.
perhaps automatic couplers and a power pack on each truck (bogie) will allow automatic shunting, which is the newtec gimmick that can make freight trains competitive again.
Haha, it seems they missed the more obvious solution: build a second story rail (with a middle rail so you can even split up or redirect workloads) and move lighter, smaller goods faster. You can now optimize for transportation of goods more valuable or time sensitive but less bulky, move them faster, and add capacity to your system, all at the same time. All on exactly the same land without interrupting existing rail. And since it's smaller and higher they could expand or reroute over properties where before it would've been too difficult.
Low tunnels and bridges would of course be an issue; some you could just go above or around, others maybe not at all. Still, there may be routes where it makes sense.
The article focuses on freight but hey there's people too. In the old days there were interurbans, and an autonomous interurban is easier than an autonomous passenger bus.
