If all trains are automated.

[Vespa]

Imagine the powers that be have managed to automate trains, would it be more efficient ?

I noticed that the Elizabeth line is automated and the train runs very frequently without any driver input, in fact I've seen a few standing up and leaning over the control to look through the front wondow with their hands over the front.

Same for London Underground, some of the line are automated with the driver opening amd closing doors.

Paris automated metro have been able to run train on an 85 seconds frequency.

Would automated mainline trains be more frequent and safer ?

 

[Sorcerer]

If the railway network was modified in a way that it could be completely grade-separated with all trains of equal speed along with the appropriate signals put in place, then I don't think there's much reason to believe they would be any less safe than automatic metros seeing as we've just essentially set it up to be a long-haul version of one in this scenario. However, the reality is that such an idea is merely a fantasy, so the question of safety is probably less relevant compared to costs and practicality.

 

[birchesgreen]

Tube trains are not fully automated, DLR more so but even that still requires staff on board to operate*.

The problem with the mainline is that you run into trouble when you integrate automated and non-automated trains (similar issue with self driving cars). In a closed system like the DLR it is easier to achieve, on the mainline you would need to convert everything which obviously has some issues.

* this depends on what you term "automated" of course, personally if the train can't operate without a staff member on board then its not fully automated.

 

[High Dyke]

Are we talking about the national rail network, with its countless level crossings, of different types?

The current digital railway project on the East Coast Main Line does suggest that frequency would increase, but there are probably a number of different factors to make that effective.

 

[whoosh]

You need platform edge doors at all platforms, which should be fun arranging them for all door positions on all the stock that uses them.

Wonder how many station overruns there'll be due to aggressive driving that automated systems love?

 

[08221]

Im in Dubai this week. The fully automated Dubai metro is incredible. So efficient, frequent and clean.

 

[hexagon789]

Im in Dubai this week. The fully automated Dubai metro is incredible. So efficient, frequent and clean.

It has broken down often enough over the years...

 

[Topological]

Why do things like crossings matter?

Surely, by the time a human realises there is a problem it is too late anyway?

An automated train would have the same signal information to make decisions from as a human driver. Assuming the route is set and the signal cleared then it must be possible for the train to cover the track in the same way one driven by a human would.

This is not saying that I want to see automated trains on all lines, but I suspect cost is the real barrier.

 

[stevieinselby]

This is not saying that I want to see automated trains on all lines, but I suspect cost is the real barrier.

In particular, the law of diminishing returns.
Where you have an intensively used system like the Lizzie Line or DLR, the cost of installing all of the equipment (including platform edge doors) is relatively small in relation to the overall operating cost of the system. As you expand onto less intensively used systems, you still need the same infrastructure in place but its costs don't significantly diminish as usage does ... but the potential for saving on driver wages or improved network performance does drop right down with usage, so you are spending the same amount of money for a vastly reduced potential benefit. The cost of fitting the wider network out for automated operation is orders of magnitude higher than the cost of continuing to run trains with 80 kilos of meat at the controls for the foreseeable future.

 

[deltic]

Imagine the powers that be have managed to automate trains, would it be more efficient ?

I noticed that the Elizabeth line is automated and the train runs very frequently without any driver input, in fact I've seen a few standing up and leaning over the control to look through the front wondow with their hands over the front.

Same for London Underground, some of the line are automated with the driver opening amd closing doors.

Paris automated metro have been able to run train on an 85 seconds frequency.

Would automated mainline trains be more frequent and safer ?

They should be more reliable as a computer can ensure journey times between stations are more consistent than humans can achieve. If unstaffed, additional services can be operated at short notice at off peak times such as for sporting events etc. in some cases stabling space can be reduced as rolling stock can be stabled on running lines.

 

[stuu]

Tube trains are not fully automated, DLR more so but even that still requires staff on board to operate*.

The problem with the mainline is that you run into trouble when you integrate automated and non-automated trains (similar issue with self driving cars). In a closed system like the DLR it is easier to achieve, on the mainline you would need to convert everything which obviously has some issues.

* this depends on what you term "automated" of course, personally if the train can't operate without a staff member on board then its not fully automated.

The DLR can theoretically run unstaffed, but TfL don't want to do that, there have been various FOI and Mayor's questions which have confirmed that.

I don't see any reason why automated trains couldn't run mixed with human-driven trains. The signalling would need to be upgraded first obviously, but it's no different to how Chiltern trains will run over the shared Metropolitan line section in principle, the Met will be ATO and the Chiltern conventionally driven.

 

[birchesgreen]

The DLR can theoretically run unstaffed, but TfL don't want to do that, there have been various FOI and Mayor's questions which have confirmed that.

I don't see any reason why automated trains couldn't run mixed with human-driven trains. The signalling would need to be upgraded first obviously, but it's no different to how Chiltern trains will run over the shared Metropolitan line section in principle, the Met will be ATO and the Chiltern conventionally driven.

There is no reason why anything can happen once you throw enough money and effort at it indeed.

 

[stuu]

There is no reason why anything can happen once you throw enough money and effort at it indeed.

True, but that's not the point, there is a real world example, in the UK, of automatic trains running alongside manual trains (or will be, I can't remember how far they have got)

Clearly it can only happen when a line is suitably resignalled, but you suggested mixed operation was an issue, when it isn't.

 

[Belperpete]

Why do things like crossings matter?

Surely, by the time a human realises there is a problem it is too late anyway?

An automated train would have the same signal information to make decisions from as a human driver. Assuming the route is set and the signal cleared then it must be possible for the train to cover the track in the same way one driven by a human would.

This is not saying that I want to see automated trains on all lines, but I suspect cost is the real barrier.

With some types of crossing, such as AOCLs, the driver has to visually check that the crossing is clear. With other types of crossing, in degraded mode, the driver is required to check for a handsignal from the person who has taken local control of the crossing before passing over it.

 

[birchesgreen]

Clearly it can only happen when a line is suitably resignalled, but you suggested mixed operation was an issue, when it isn't.

Well it is an issue if you have to resignal a line before it can take place surely? Maybe we have different definitions of what "issue" means? To me it means something you can't do immediately without changes of some sort.

 

[Belperpete]

Automatic operation gives consistent train performance, which is a real benefit on a high-density traffic line, as it means all trains behave in the same way. With multiple trains following one another, performance is only as good as the weakest link in the chain. On a mixed traffic railway, and particularly one with junctions that significantly degrade the throughput, automation gives much less benefit. It was why the core section of the Lizzy line was automated, but not the bits outside.

Once the Victoria line had been semi-automated, it was seen that if you were going to have a person sitting at the front of the train, then they might as well be driving it. Particularly as LT ended up paying the ATOs more than ordinary drivers!