Moving Block signalling - when will it reach mainline railways?

[BRX]

As I understand it, the only places Moving Block signalling has been implemented so far are essentially metro lines.

I think this is partly due to the fact they are less complex and often have few junctions?

Is it close to being implemented anywhere in a more main-line context yet? How many years away is it, before it starts to become widespread?

 

[zwk500]

As I understand it, the only places Moving Block signalling has been implemented so far are essentially metro lines.

I think this is partly due to the fact they are less complex and often have few junctions?

I suspect a key factor is that the trains are all of the same type, so the braking distance maths only has to be done for one unit. It's also because the capacity is generally more valuable on metros than mainline railways

Is it close to being implemented anywhere in a more main-line context yet?

No. Certainly not in the UK. Most computer systems can't even simulate it properly. It's part of ETCS Level 3 specification, but no railway in Europe has a firm implementation date for any line at that level, I think. The UK has an additional problem that our infrastructure data is nowhere near what it would need to be for Moving block.

How many years away is it, before it starts to become widespread?

Many. Although it's very difficult to put a precise number on it because if computers continue to improve at current rates (especially with Quantum computing etc) then the maths behind it could become a lot quicker to process.

 

[coxxy]

we still have a fair few lines using Absolute block and Semaphore signalling, old school signal boxes and too many level crossings.. At a guess, We are generations away..

 

[Sonik]

Is true moving block even necessary?

An advantage of ETCS is that Ballises are much cheaper to deploy than fixed train detection (i.e. track circuits/axle counters). This allows much shorter blocks to be deployed economically, with the number of blocks required for headway varied depending on speed.

The key enabler, is the equipping of trains to allow this to work, particularly with regard to confirming train integrity of freight consists. Was discussed at length on another thread, I will see if I can find it.

 

[D365]

Is true moving block even necessary?

An advantage of ETCS is that Ballises are much cheaper to deploy than fixed train detection (i.e. track circuits/axle counters). This allows much shorter blocks to be deployed economically, with the number of blocks required for headway varied depending on speed.

The key enabler, is the equipping of trains to allow this to work, particularly with regard to confirming train integrity of freight consists. Was discussed at length on another thread, I will see if I can find it.

Agree with all these points. On a mixed-traffic railway, there simply isn't any need for moving block. If you'll forgive the turn of phrase - much lower hanging fruit to be picked first.

 

[BRX]

Is true moving block even necessary?

An advantage of ETCS is that Ballises are much cheaper to deploy than fixed train detection (i.e. track circuits/axle counters). This allows much shorter blocks to be deployed economically, with the number of blocks required for headway varied depending on speed.

The key enabler, is the equipping of trains to allow this to work, particularly with regard to confirming train integrity of freight consists. Was discussed at length on another thread, I will see if I can find it.

Is this what the ECML is getting (more, shorter blocks) or do the blocks remain as they are now?

 

[zwk500]

Is this what the ECML is getting (more, shorter blocks) or do the blocks remain as they are now?

The marker boards won't be exactly where the signals currently are on the Signals-free section. They're likely to still be a similar distance apart on plain track, but to get shorter around complex junctions (every block is another piece of interlocking code, and developer's time isn't cheap to purchase). I think a headway improvement is part of the plan though. Where it's L2 Overlay, the Signals will be where the block markers are (I don't know if there will be actual marker boards on the ground or not). I'm open to correction, as it's been a while since I spoke to anybody involved with the plan.

 

[Railsigns]

Is true moving block even necessary?

An advantage of ETCS is that Ballises are much cheaper to deploy than fixed train detection (i.e. track circuits/axle counters). This allows much shorter blocks to be deployed economically, with the number of blocks required for headway varied depending on speed.

With ETCS, you can only dispense with continuous train detection if you go for Level 3, and you're going to fit Level 3 you may as well have moving block, which no other ETCS level offers.

 

[HSTEd]

With ETCS, you can only dispense with continuous train detection if you go for Level 3, and you're going to fit Level 3 you may as well have moving block, which no other ETCS level offers.

The two systems become effectively indistinguishable from each other as the block length decreases.

Once you reach blocks that are significantly shorter than the lengths of the trains (say 100m or less) there is little point with the conceptually and technically far more challenging moving block.

But in any case, it took years of negotiation to get freight operators to agree to ETCS.
We have no chance of persuading them to willingly adopt electronic train completeness check equipment, even though it exists, and no political appetite to force the issue.

 

[zwk500]

The two systems become effectively indistinguishable from each other as the block length decreases.

Once you reach blocks that are significantly shorter than the lengths of the trains (say 100m or less) there is little point with the conceptually and technically far more challenging moving block.

Agree. Level 3 would only really be worth it in dense metro operations, like London Overground or Castlefield, where every last metre really does matter.

 

[HSTEd]

Agree. Level 3 would only really be worth it in dense metro operations, like London Overground or Castlefield, where every last metre really does matter.

I think the Level 3-associated on board train detection is potentially useful, but not the moving block technique. So essentially the L2/L3 Hybrid system the working group has been talking about for the last few years.

 

[TSG]

Many. Although it's very difficult to put a precise number on it because if computers continue to improve at current rates (especially with Quantum computing etc) then the maths behind it could become a lot quicker to process.

Perhaps you know different for a fact, but I really struggle to believe that computing power is a limiting factor on this (even allowing for the use of less than cutting edge processors for safety applications and software written to optimise verification rather than performance), Processor systems in aerospace and defence already achieve more than is required for moving block.

Surely the problem isn't doing the calculation but proving that you are putting the right numbers in to the algorithm for each train, for all the infrastructure (as you allude to) and for the conditions, every time (whilst avoiding the use of such large safety factors that you end with less capacity rather than more). I'd also imagine dependable communications bandwidth would be more of a technology limitation than the number crunching.