Melbourne needs high-capacity signalling

Melbourne's train network is at a tipping point and will struggle to cope within a few years unless high-capacity signalling is introduced, the state government's transport boss has warned.

Public Transport Victoria chairman Ian Dobbs said there were "a whole lot of different pressures that will emerge on the rail system in the next few years" that, without the necessary investment, would result in passengers being increasingly left behind on stations at peak periods.

High-capacity signalling, which tells train drivers when the line ahead is clear by putting the signals inside their carriage instead of alongside the railway tracks, would allow trains to be run more efficiently across the network. But while the technology is used in dozens of cities around the world, it has not yet been introduced in Australia.

Quote:

Melbourne's train network is at a tipping point and will struggle to cope within a few years unless high-capacity signalling is introduced, the state government's transport boss has warned.

Public Transport Victoria chairman Ian Dobbs said there were "a whole lot of different pressures that will emerge on the rail system in the next few years" that, without the necessary investment, would result in passengers being increasingly left behind on stations at peak periods.

High-capacity signalling, which tells train drivers when the line ahead is clear by putting the signals inside their carriage instead of alongside the railway tracks, would allow trains to be run more efficiently across the network. But while the technology is used in dozens of cities around the world, it has not yet been introduced in Australia.

Full story here.

I don't know what kind of high-capacity signalling is being referred to here. Can someone enlighten me, at least to the type used in the UK? Maybe it is known by a term more familiar to us. Thanks.

modest bosting capacity from 20-25 tph on high density coded track circuits to maybe 30 or 35tph

I think everywhere has to be in radio communication

ETCS Level 2 does, else how are movement authorities transmitted to trains? Level 1 uses the balises to transmit movement authority but this means they have to be remotely controlled (as opposed to just powered from trains), so you may as well supplement them with lineside signals when you're doing that.


With good modelling of local terrain, etc. you can get a pretty reliable measure of radio coverage. Of course there is the problem of possible interference, but as long as you have good regulation to ensure nobody else is going to be using that band, only freak cases of natural interference (most of which only tend to affect lower frequencies than that used for GSM-R anyway) are of any consequence. I would be very surprised if portions of the GSM-R fitted railway network in this country have poor GSM-R coverage, and this isn't considered a safety problem.

These numbers seem quite low considering that Prague metro is designed for 45 tph and uses plain old coded circuits, not really sure if it could be called high density (maybe). Also, 20tph should be possible to carry over classic lines, eg. fast lines north from Cross - it's not really that much.

I'm not sure that's true that transmitting through the rails would be cheaper, I don't recall hearing anyone complaining about the GSM-R budget for fitting the substantial portion of the network that now has it. I would suspect you need a lot less kit for radio, but even if this isn't true I believe there must be a good reason why it was chosen for ETCS. I'd like to know what that reason is if anyone knows!

While the VLine Regionals, and the Broad Gauge Freight share the network with Metro, they won't be able to go with Metro type signalling.

" said:

I'm not sure that's true that transmitting through the rails would be cheaper, I don't recall hearing anyone complaining about the GSM-R budget for fitting the substantial portion of the network that now has it. I would suspect you need a lot less kit for radio, but even if this isn't true I believe there must be a good reason why it was chosen for ETCS. I'd like to know what that reason is if anyone knows!

Because consultants and signal vendors convinced railway management that it would be cheaper. Also Germany seems to be scared of track circuits for some reason (and I think the UK caught their flu).

That's what happened with all the Positive Train Control junk in this country. The initial plans called for a smaller number of transmitting stations and the railroads were hot to avoid having to install transponders and other physical components. However when it started to get installed radio coverage became an issue so at the end of the day the cost conscious freight railroads wound up just as much kit as the "higher footprint" transponder and track circuit system that was already in service on parts of Amtrak. The wireless systems are still not in service in any meaningful way and those few that have had bandwidth issues along with costly frequency changes.

Non-technical management always assumes that "wireless" means lower cost. That's true if your setup can tolerate 8's of service and/or poor security, but if you want high availability stick with hard data links. New Jersey Transit has been the one outfit to not buy into the BS and go with a system that uses absolutely no data radios. The ACSES system on Amtrak can actually still function if the data radio component fails as well. Both use coded track circuits for most of the safety critical functionality.

BTW didn't that pilot ETCS system in Wales cost something like 400 million pounds for a single line?

The £400m cost apparently includes a lot of research and development that would apply equally to other lines in the UK so does not need to be repeated. I mean, it's a pilot scheme using new technology sparsely implemented elsewhere, what honestly do you expect?

It would surprise me if that £400m had a significant portion to do with GSM-R. Apparently the whole south of England (except the South East which was previously fitted with the newer CSR system instead of NRN, but this will eventually be replaced too) is now fitted with GSM-R, as well as the two most important main lines (West Coast and East Coast), and again, I've not heard about any significant cost complaints. Just because it might be unfeasible in the sparsely-populated and bloody huge US doesn't mean it's not feasible here.

The Boston Green Line trolley system actually runs at such a short frequency it exceeds the capacity of all CBTC systems. At peak periods the block signals are generally taken as "advisory" and the system is run on sight. The unions actually like this because they can implement rulebook slowdowns when the contract is being negotiated. Anyway, with the way autonomous vehicle technology, various forms of adaptive cruise control might be a better solution than CTBC if you have high enough braking rates. With computer vision the rail vehicle can detect speed restrictions the same way a human can.

Also Germany seems to be scared of track circuits for some reason (and I think the UK caught their flu).

Even today CSX is ripping out its ATCS packet radio* pole line replacement because they got a better deal on a satellite communications package.

*ATCS messages are authenticated so one can't just flip a switch and turn SimSig into RealSig.

Wait... so the rule book is used when it's quiet or the drivers want a pay rise, but thrown out the window if it's busy?! :doh

Supposedly a lot more reliable to use axle counters but I'm not entirely convinced. Possibly slightly lower maintenance because you don't have to fiddle with resistance and variable track conditions. But it massively complicates the interlocking logic due to the ability for a signaller to clear a failure under certain controlled circumstances.

Most people whinge about detection of broken rails but it turns out - according to US railroads - relatively few breaks are detected in this way. You need a clean pull-apart and (I forget the exact figure quoted) it's way under 50%.

I have an antennae sitting on my roof receiving ATCS packets so I've done a little digging. Firstly there is no authentication, not in the sense of "did the control center just send me that message", just a kind of checksum checking. I don't want to go into specifics on an open forum but given the right equipment messages can be sent to these interlockings easily enough. The saving grace is that (a) the interlocking is done locally so anything it receives is only a request, not a command so it won't do anything unsafe; and (b) the control centere would (eventually) notice that its commands and indications were out of sync.

" said:

Most people whinge about detection of broken rails but it turns out - according to US railroads - relatively few breaks are detected in this way. You need a clean pull-apart and (I forget the exact figure quoted) it's way under 50%.

Remember it is fairly difficult to derail a train so those breaks most worth detecting might only be the large pull aparts.

" said:

I think everywhere has to be in radio communication

That sounds like a de-facto violation of physical laws.

Quote:

Melbourne's train network is at a tipping point and will struggle to cope within a few years unless high-capacity signalling is introduced, the state government's transport boss has warned.

Public Transport Victoria chairman Ian Dobbs said there were "a whole lot of different pressures that will emerge on the rail system in the next few years" that, without the necessary investment, would result in passengers being increasingly left behind on stations at peak periods.

High-capacity signalling, which tells train drivers when the line ahead is clear by putting the signals inside their carriage instead of alongside the railway tracks, would allow trains to be run more efficiently across the network. But while the technology is used in dozens of cities around the world, it has not yet been introduced in Australia.

Full story here.

I don't know what kind of high-capacity signalling is being referred to here. Can someone enlighten me, at least to the type used in the UK? Maybe it is known by a term more familiar to us. Thanks.

We have true moving block over here on the Docklands Light Railway, Jubilee Line and Northern Line. I believe all these use 'wiggly wire' transmission. All seem to work remarkably well; it took a while to beat the bugs out of the Jubilee Line but the Northern Line went in very smoothly indeed. The sub-surface lines (Metropolitan/Circle/District/Hammersmith & City) are down to be done next, using radio transmission rather than wiggly wire, but as the whole UndergrounD network is covered by leaky-feeder radio already (and, I believe, with very high reliability) there shouldn't be too much of a problem.