We had a major crash in Australia some years ago, when a tourist steam train deposited sand on the track of an incline to improve grip this sand turned to glass and essentially insulated to steam train from the track and did not show up on the Circuit signal

The
challenge today is to replace the venerable (but coarse grained and
high maintenance) track circuit with GPS and other train location
technologies. This can provide a near-real-time train location
capability rather than a block-to-block one, which can potentially
increase capacity by reducing train spacing. Because a track circuit can
detect when a train has vacated the circuit, such alternative location
technologies would also need to know where the end of the train is
located.

One of the challenges is that track circuits can give
some indication - but not a perfect one - that a rail has broken (due to
temperature-based contraction or due to fractures and other
imperfections failing under repeated impact). A separate
technology-based solution for this is therefore required to fully
replace track circuits, and hopefully a more robust one than the track
circuit, since many kinds of broken rail events (let alone expansion
kinks that are just as dangerous as broken rails) do not interrupt the
current flowing through the track circuit.

Hi Thomas & Bill,
Re the accident in Australia,
I was working as a volunteer with another steam preservation society at the time, and remember it well.
The sand issue was complicated by the fact that the train had stopped on a very heavy grade, 2.5% or 1 in 40, and had to set back a few feet to re-start, allowing the engine's front truck wheels to roll back onto the sand that had been applied in front ot the driving wheels.
Even though momentary, the insulation of the whole train on sand cleared the signal behind and a following train proceeded.
It always seemed unacceptable to me that a signal system could allow a train to "disappear" and clear a signal, without having been detected in the section ahead.
But as far as I know that was not one of the recommendations of the coronial enquiry.

Even with all of the GPS, radio and virtual moving block technology, I think the traditional automatic block signal and the manual interlock signal are the best and most fail-safe train control components that there are.

All of the cautions reported are true. The most worrisome problems t o me are 1. Trains derailing off main double track give no warning to oncoming trains on the opposite track. 2. At sidings with manual switch points use D-rails to protect. D-rsils take the train off of the track circuit, yet I have seen wagons roll out onto the tracks, "on the ground" yet still off the signal system, giving no indication to the signals. D-rails are a Track department invention that does not protect anything!

If you REALLY want to know more about the history and operation of railroad signals, go to epri.com and search for document number 1012652. You can download it for free. Chapter 4 is 160 pages of how signals work.

My father-in-law lived much of this history during the late 19th Century. I loved listening to his stories about the rail roads.

@Larry: In addition to my current status with the IEEE, I was a US-licensed Locomotive engineer for about 10 years in the past. Even on inter- and intra- State rails, there are some differences in signaling, and US regulations require a minimum degree of familiarity before any engineer is left to operate any section of rail on his/her own. On some occasions, rail vehicles needed to travel between the various Amtrak(national) / State / and local commuter rails.. our orders generally were to use an experienced engineer ( in the locomotive-operator sense of the word) , who knew the territory as a pilot, while the operator who was experienced with the equipment actually handled the vehicle operation.
This is a lot like how freight ships are required to operate in harbors - a local harbor master or tugboat operator will lead/instruct as a pilot/navigator, by radio or onboard, while the vessel captain actually operates the mechanical aspect.
The pilot, in addition to knowing the "lay of the land" will also be familiar with communication requirements, traffic patterns, and does most of the talking with controllers ( rail or maritime).
Also, from my travels with Amtrak, and various PBS travel shows, the general process whenever a train changes "zones" is for a crew change and a locomotive change to take place. These zones can be mechanical ( electrified rail, vs overhead electric, or diesel) or at State/National borders - with exceptions in some cases like the "Chunnel" rail line, where the train crew only does that specific International (but short) run.
In the US, this also happens when trains leave the mostly-electrified Northeast lines, and changeover to Diesel ( i.e. Washington DC: Amtrak locomotives heading North-South are changed out here. Diesels handle southern passenger lines, electrics go north) .
In Europe, it seems to take place at most national crossings, at least from the 3 or 4 travel shows that actually showed the process.
Prior to 1990-ish, I'm not sure what arrangements were made. I can check with some friends who are rail historians or "rail buffs" if anyone is really curious.

@c.k. - amen to that. Signals and communication affect nearly every aspect of life; having even a simple set of mutually-understood rules is close to miraculous.

@Thomas Hadley: Wow! This should be taught as a case-study for engineers. I rarely ran lines that ever needed sand or other "traction adjusters", but we were repeatedly cautioned not to over-sand; and advised when sand had been applied on the rail in front of us. Do you know if any special steps are taken to avoid this happening again?

@Marco - unfortunately, that is not limited to the UK. It is a deprecating joke in the NYC metro area as well; nearly all the commuter rail lines in the 100 or-so mile radius around NYC suffer the same fate almost daily. As an insider, I do know that simple phrases like "signal trouble" are over used, but a lot easier to communicate than " the train had a square wheel which wasn't found on inspection", " we intentionally stuffed two trains on a single block - so the safety overrides were cut out, and won't go back in place", or " the railroad is so old that heat/cold has made thermal expansion a big issue, and a switch / bridge / safety circuit won't properly clear".. and the hundreds of other reasons that could delay a train. The basic rule is if one train gets delayed, every train behind it is also delayed. During heavy travel, this delay gets magnified- and won't be "reset" until the problem is solved, AND all the traffic for that particular "rush" is cleared.

I wonder when trains fornthe first time left their country, what happen then? With hundreds of differnt signalling standards then and still dozens today I wonder how directing a train works at all.

Classic history and relevant for any engineer interested in the theory and practice of signaling and communication. Should be taught in school so kids learn the terrible challenges involved in communicating successfully between human beings - despite speaking the same language.

Classic history and relevant for any engineer interesting in signaling history and the human predicament.

The invention was good but the implementation not so resilient: in UK the most repeated and heard every day announcement in all the stations is : "the train has been delayed for a signal failure..." :-)