Autonomous Rail Transportation

JonMyrlennBailey

Active member
If there ever was a case for autonomous transportation, the rail industry takes the cake. Trains are guided by tracks and directed by junctions. All this can be computerized and unmanned. There is no reason under the sun for any freight railway employee these days to not come home each and every work night to sleep in his own bed. A freight train could be sent from California to New York totally unmanned. Furthermore, long-haul motor freight, commercial trucks, on highways should be minimized and long-haul rail freight should be maximized for peak freight efficiency. How well could Dave, Geoff or Bald drive a physical train?

Here is one example of an unmanned freight down under:

(15) Rio Tinto Autonomous Train - YouTube



EXTRA! Modern-day freight train robberies totally trashing the American landscape:

(15) Thieves raiding rail cargo containers in Los Angeles - YouTube

Question: how can railroads curb messes like this?
 
Last edited:
I'm not 100% against auto-trains, but I look at other industry's for context and concerns. An airbus jet can go from the start of the takeoff runway to stopping on the landing runway on full auto pilot, just needing the crew to punch in all the data, witch could be done remotely. I've also seen ideas and test floated for auto taxi form gate to runway. However its still very much industry standard (and to my knowledge the only people wanting it to not be are in the accounting departments) to have 2 trained pilots at the front to act in case's of "oh lord the plane is on fire" ect. I do not see the engineer getting booted out of the cab on class 1's in the US, conductors are at risk constantly but Engineers no.

For 1, unions would riot. Unions have pros and cons, and I don't care much for that in this discussion but, yea. they would riot. For 2, trains hit Darwin award candidates every day with near miss's 4x fold. With full auto no one in the cab or remote controlling it I doubt an auto system could see far enough to avoid an avoidable accident, the ones where the "target" get stuck on the tracks a decent time before the train is on top of them and the crew is able to spot it and stop. Even more so crews can avoid impacts using the horn and watching while predicting what someone is going to do to act in a stupid way, like when you going down the highway and, after years of watching, have a sixth sense about a car in front of you about to do something stupid. Teaching a non self aware AI that is nigh impossible.

For 3, with all that said, do you think your local town would support this? I live near Chicago and "idiots" are constantly wailing about how noisy and annoying and dangerous trains are and why do they have to go though our lovely (very expensive to live in) suburb and couldn't they just not etc. etc. etc. Do you honestly think that people like that would go along with it? For 4, they would make fairly easy targets for people wanting to cause havoc. Breaking into the cab/ hoping on a loco and destroying it, hacking the thing and making it do very bad things. You have to make things idiot proof, and these ant it. at least for a while in the US.

The reason robo-trains are a thing in Australia is most of them (to my knowledge) are in the dessert, barely anything around, hauling ore. They are only around people at the mine, where its ostensibly only workers who know how to act around them, or at the port town, where I'm sure they are gated off quite heavily. In this context of going back and forth hour after hour day after day in an area where the only things to hit are wildlife (that probably isn't going to make a human crew slow down either) it works. in the US where it eventual will have to pass though an area with people in it, nah.

Looking back on this I got a bit ranty, sorry about that, I've had this debate in my head more than once :wave:
 
Well said Folks,

There is much to be considered, and AI can not do it all, that as said 6th sense in one thing us Humans have if we only listen to our Gut Instinct, and I have survived OK with my Gut instinct, I cannot write the details here, but have had some very close calls on Situations, and will leave at that!

I just noticed in my area that UPRR has closed off one of their Mainline Tracks coming thru my area, and it looks to me like now the Trains have to back into a certain Rail Yard, rather than pull in, like a Stub Yard.......I can't for the life of me understand why they did this? :hehe:

And in fact because part of the Main has been removed, one of their 2mile long trains stopped in my area the other day for several hours, and blocked Crossings, hmmm, could that be because there was no alternate route open one mile to the North, because the main is now stubbed out to the Rail Yard.....

This is supposed to be next level Precision Railroad Scheduling, I think it, like "JUST IN TIME INVENTORY" isn't exactly working out, just like we are Short Gasoline do to the War.......:eek:

Excuse my Rant, I need to go stand in the Corner.......:(
 
Auto trains could have radars just like Toyota cars with safety sense. They can have GPS. They can have cameras, heat detectors and motion detectors to detect trouble at crossings and ahead on the tracks. The AI computers would be tied in with railroad signaling. Much of the local railroad work as in the yard could still be done by hand. There is no reason to put humans inside long-haul freight trains anymore. Sitting on board a train for long hours and days away from home is not a healthy lifestyle just as OTR truck driving is not healthy. Manned freight vehicle operation on road or rails can be limited to short haul.
 
Last edited:
Auto trains could have radars just like Toyota cars with safety sense. They can have GPS. They can have cameras, heat detectors and motion detectors to detect trouble at crossings and ahead on the tracks. The AI computers would be tied in with railroad signaling. Much of the local railroad work as in the yard could still be done by hand. There is no reason to put humans inside long-haul freight trains anymore. Sitting on board a train for long hours and days away from home is not a healthy lifestyle just as OTR truck driving is not healthy. Manned freight vehicle operation on road or rails can be limited to short haul.


You are simply incorrect. It sounds good when viewed as a general idea, but it is practically impossible. All of the things you have listed are already a large part of railroading; GPS, cameras, IR detectors and radar have all been used for years. The latest signalling and control systems available still don't even work with a fraction of the reliability that would be required to fully automate even a single train. It may happen, but is still many years ahead.
 
Auto trains could have radars just like Toyota cars with safety sense. They can have GPS. They can have cameras, heat detectors and motion detectors to detect trouble at crossings and ahead on the tracks. The AI computers would be tied in with railroad signaling. Much of the local railroad work as in the yard could still be done by hand. There is no reason to put humans inside long-haul freight trains anymore. Sitting on board a train for long hours and days away from home is not a healthy lifestyle just as OTR truck driving is not healthy. Manned freight vehicle operation on road or rails can be limited to short haul.

The biggest problem with all of that is range. IR, laser detection, radar, they all have pitiful effective ranges when your near the ground due to ground clutter/ the system not being able to tell what "object 7492" is. That (sort of) works in a car, where the breaking distance is feet. I would check out people testing the "improved" tesla auto drive and how many times that systems try's to kill them/ anyone around them because it as no idea what its looking at.

Now consider a 15000ft consist rolling at 70mph in the flatlands on Kansas, and a truck has gotten stuck on a crossing ahead. A (admittedly VERY) sharp eyed crew could spot that, and start slowing, maybe not stop in time, situation depending, but certainly lowering the energy of impact, witch helps lower the damage to the locomotive/ potential for derailment and lessens the rick of flying debris. The truck is still scrap, of course, unless the consist can stop in time but low impact energy is always better. Even at slow speeds these systems would struggle to see dangers ahead.
 
The biggest problem with all of that is range. IR, laser detection, radar, they all have pitiful effective ranges when your near the ground due to ground clutter/ the system not being able to tell what "object 7492" is. That (sort of) works in a car, where the breaking distance is feet. I would check out people testing the "improved" tesla auto drive and how many times that systems try's to kill them/ anyone around them because it as no idea what its looking at.

Now consider a 15000ft consist rolling at 70mph in the flatlands on Kansas, and a truck has gotten stuck on a crossing ahead. A (admittedly VERY) sharp eyed crew could spot that, and start slowing, maybe not stop in time, situation depending, but certainly lowering the energy of impact, witch helps lower the damage to the locomotive/ potential for derailment and lessens the rick of flying debris. The truck is still scrap, of course, unless the consist can stop in time but low impact energy is always better. Even at slow speeds these systems would struggle to see dangers ahead.

Then, they could put a camera in the lead engine to look down the track. At some stationary building, a human watching a TV screen could monitor the track ahead of the train and command the train to stop remotely in case of trouble spotted. I bet over the next ten years, long-haul mainline (Class 1 in train speak??) freight trains over American soil will be, for the most part, unmanned. We can all agree that trains can be more easily automated than cars and trucks over highways. After the eight hour shift, another human can come to the RR office and continue to monitor the trains. As a matter of fact, one might have eight TV screens being monitored by one human at the same time, one screen for one train. Each TV set could have a red button to punch in case of emergency to stop the train corresponding to it, say a moose on the track or a truck that crapped out at a crossing. This minimizes payroll costs of railroads. Punching the red panic button would put the train into full emergency stop mode. I'm sure as time passes, the industry will have all this rocket science advanced. The automobile and the airplane did not really get practical, comfortable, affordable and safe until several decades after its invention.
 
Last edited:
Then, they could put a camera in the lead engine to look down the track. At some stationary building, a human watching a TV screen could monitor the track ahead of the train and command the train to stop remotely in case of trouble spotted. I bet over the next ten years, long-haul mainline (Class 1 in train speak??) freight trains over American soil will be, for the most part, unmanned. We can all agree that trains can be more easily automated than cars and trucks over highways. After the eight hour shift, another human can come to the RR office and continue to monitor the trains. As a matter of fact, one might have eight TV screens being monitored by one human at the same time, one screen for one train. Each TV set could have a red button to punch in case of emergency to stop the train corresponding to it, say a moose on the track or a truck that crapped out at a crossing. This minimizes payroll costs of railroads. Punching the red panic button would put the train into full emergency stop mode. I'm sure as time passes, the industry will have all this rocket science advanced. The automobile and the airplane did not really get practical, comfortable, affordable and safe until several decades after its invention.

Yeahhhhhhhhh, that unlucky person is gonna get real sleepy real fast and will fall asleep. Who's gonna watch over the trains then? What about when he gets bored of watching a TV for six hours? Will he be asleep? Will he go somewhere and get a snack? Who's going to watch the train while he's gone?
 
I don't know, but I don't think I could get paid enough money to sit in an engine cab long-haul for a living. Those train crews that deliver and pickup freight cars locally to and from customers probably sleep in their own bed each and every night.
 
I get what you are saying, but even the local folks can get called any time of any day or night to come out and put together a train that has to be delivered somewhere. Their hours are really unpredictable, and it can really throw a monkey wrench into plans and family life. I think I might rather be the long-haul guy; I never get tired of scenery. Watching it on a TV? Not so much, but THAT guy can maybe go home when the shift relief comes, and he can go home and sleep in his bed. I would want a LOT of cameras to look around that train with though.
 
Last edited:
Local rail freight service is unpredictable? During the 1960's, 1970's and 1980's, the local SP pickup/delivery trains came through my northern California coastal county town of Novato like clockwork. They more or less ran like local postmen in mail trucks. Maybe some RR's run on a tighter schedule than others. California is heavily union too. Union bosses don't want their dues-paying members working crappy odd hours.
 
AI control of vehicles -- rail or road -- just isn't advanced enough to handle 100% of all driving situations. AI can handle 90-95% of it currently, but hasn't been able to close the gap. As bnsfc mentioned, the Australian robot trains are isolated from the general public.

I like the idea of two man crews on trains. Even if the engineer is the only one at the controls, the conductor can still keep an eye on the train and on the engineer. I believe that sitting at a desk would be even harder than sitting in a train when it comes to boredom and alertness.

Locomotive and rail car maintenance would have to be improved to prevent locomotive and rail car failures.

The automation ideas floating about sound good, but still aren't quite practical.
 
AI control of vehicles -- rail or road -- just isn't advanced enough to handle 100% of all driving situations. AI can handle 90-95% of it currently, but hasn't been able to close the gap. As bnsfc mentioned, the Australian robot trains are isolated from the general public.

I like the idea of two man crews on trains. Even if the engineer is the only one at the controls, the conductor can still keep an eye on the train and on the engineer. I believe that sitting at a desk would be even harder than sitting in a train when it comes to boredom and alertness.

Locomotive and rail car maintenance would have to be improved to prevent locomotive and rail car failures.

The automation ideas floating about sound good, but still aren't quite practical.

AI rail service will someday become a grave necessity when humans no longer want to drive trains for a living. The American trucking industry is also short about 50,000 drivers nationally because nobody wants to be cooped up in a rig away from home. Long-haul manned vehicle operation for a living is a very unhealthy lifestyle. I still think AI is now feasible enough to send unmanned freight trains long haul as between two yards hundreds or thousands of miles apart. Manned trains might then be limited to local work as customer pickup and delivery. If trains break down, drivers aren't mechanics anyway. In this day of unmanned spaceships to Mars, there is no excuse for unreliable rolling stock on lowly earthbound railways. Much of human RR work then would be keeping the vehicular components of trains as well as train tracks in ship shape. If locos and rail cars tend to crap out a lot, maybe then Toyota should build them.
 
Last edited:
... In this day of unmanned spaceships to Mars, there is no excuse for unreliable rolling stock on lowly earthbound railways. Much of human RR work then would be keeping the vehicular components of trains as well as train tracks in ship shape. If locos and rail cars tend to crap out a lot, maybe then Toyota should build them.

It seems to me that you are comparing apples to oranges. There is nothing cheap about the incredible technology used in space travel. They spend millions of dollars to transport a ton of goods to Mars.

I'm not saying that locomotives and rail cars are cheap, but railroads try very hard to spend the least amount of money possible to transport a ton of goods.

I don't know what practices are like today, but I read that in 2000-2005 that the train crews were the ones that repaired broken coupler knuckles themselves with a spare that they carry on the locomotive.

Automation has nearly replaced the engineer. As far as I know, every modern loco manufacturer includes software that can record and learn the throttle settings of a specific route to maximize fuel savings. I don't know how well the software can adjust to unusually poor traction conditions or equipment failures.

There are many jobs that are not pleasant to perform. That really isn't a valid reason for getting rid of a job. I don't think that it is unreasonable when moving 10,000 tons of freight from point A to point B that there should be a couple of people on board to keep an eye on things.
 
The biggest problem in this thread and others is that all of your positions or ideas are based on wild assumptions that are simply wrong.

AI trains will be a thing when the technology allows it to be better suited to driving the train than a human. This has not occurred nor is anywhere near practical at this time.
 
The biggest problem in this thread and others is that all of your positions or ideas are based on wild assumptions that are simply wrong.

AI trains will be a thing when the technology allows it to be better suited to driving the train than a human. This has not occurred nor is anywhere near practical at this time.

I beg to disagree. I am a visionary. This is how I see tomorrow today. My 1908-born grandfather swore up and down that man would never go to the moon. I don't want to be a naysayer. What boomers here would have dreamed of smartphones in their lifetimes? I never claimed that this AI thing is perfect right now. It's not even perfect yet in the Trainz game. I predict that AI freight hauling by rail will become commonplace in America before I die of old age. I'm sure interested in following its progress. It's not a wild assumption at all. We all know AI rail transportation is already happening somewhere in this world today.
 
Last edited:
I beg to disagree. I am a visionary.

elissa-slater-big-brother.gif


If there's one thing I can hand it to you it is how you're able to be so confidently correct and wrong at the same time.
 
elissa-slater-big-brother.gif


If there's one thing I can hand it to you it is how you're able to be so confidently correct and wrong at the same time.

So, I am seeing the future wrong? Ok, I'm only making predictions here. Only time will tell what is correct and what isn't about what is to come.
As I have said, unmanned trains are already happening somewhere in this world today with these videos as my witnesses:

https://www.youtube.com/watch?v=fSMl1O9xxp4
https://www.youtube.com/watch?v=a52ZSCIAoRk
https://www.youtube.com/watch?v=Fyeb8AQig3w

Of course, many here frown upon the notion of real-world driverless trains. You see it as future human job loss, no doubt. America will probably be one of the last countries to jump on board this "auto train" tech due to long-standing unionization of our railroads.
 
Last edited:
I think robotic trains are even more feasible than flying cars for now. By not having train crews on board freight trains, it saves lives in case of train wrecks. It saves human health too. Have you seen those obese truck drivers? Loco driving long haul must be even more sedentary. They can build train couplers out of more durable materials. Spread locos evenly out over the length of a long freight train and coupler stress is minimized along with derailment probability. They can build Rolls-Royce-grade locomotives and rail cars. They at the rail industry don't have to cheap out when laying and upkeeping train tracks.
 
Last edited:
Back
Top