How reliant are UK railways on the Global Positioning System?

[Horizon22]

Automatic Selective Door Operation (I believe) - so that trains longer than the platform open only the appropriate doors.

Yes I believe this would probably be one of the biggest issues. Many long trains would presumably fail to open the doors as they should at stations with short platforms. It could be overridden locally, but would probably lead to delays.

 

[djw]

Active distortion of positions can be easily combatted using DGPS (differential Global Positioning System) technology - basically using GPS to plot the position of a fixed ground station, and then transmitting the position error to surrounding receivers so they can apply the same adjustment to their position. As this effectively renders selective availability pointless, the US has 'promised' that they will never again use it on GPS.
I don't think any nation would be likely to start to shut down systems that are absolutely safety critical to commercial aviation, and are virtually safety critical to merchant shipping as well, though ships do have more alternatives, especially in coastal waters, without giving at least enough notice to get aircraft out of the sky. It is difficult to envisage a Domesday scenario which requires all 3 systems to be shut down instantly - perhaps an indiscrimate firing of nuclear missiles by China, but I can't imagine they rely totally on foreign satellite navigation systems anyway.

Terrestrial DGPS is old school now. The UK and Ireland marine DGPS is being shut down by the General Lighthouse Authorities on 31 March 2022 - the equipment is life-expired and the accuracy available from GPS alone (without the benefit of the GLONASS, BeiDou and Galileo satellites used by the more modern multi-constellation GNSS receivers) typically meets the DGPS performance specification most of the time now that Selective Availability has been permanently discontinued. Integrity monitoring can be done with modern RAIM (receiver autonomous integrity monitoring) capable receivers - it doesn't need DGPS.

The newer approaches to augmentation are space-based augmentation services (SBAS), such as EGNOS in Europe, or assisted GPS, where the augmentation is retrieved via the Internet.

Selective Availability will not come back on GPS. There is no Selective Availability hardware fitted from Block IIF satellites onwards, so of the 31 satellites in the current GPS constellation, only 14 have SA hardware. This number will decrease over time as the older Block IIR and IIR-M satellites are retired. Moreover, it would be pointless for the United States to start intentionally degrading civilian L1 GPS signals again when there is no such thing as Selective Availability on GLONASS, BeiDou, Galileo or the L2C and L5 signals of GPS.

Unsurprsingly, other people have noticed that over-reliance on GPS is a risk. LORAN may be on its way back as eLorain: https://www.xyht.com/gnsslocation-tech/the-return-of-loran/

Northwestern Europe had eLORAN - but France and Norway decided to turn off their transmitters on 31 December 2015. Without the French and Norwegian stations, the LORAN chains covering UK and Ireland would not work, so the General Lighthouse Authorities terminated the UK and Ireland stations.

Ongoing thought is being given to GNSS vulnerabilities as the satellite signals are weak and easy to jam, whilst the system owners seem determined to limit the use of the encrypted and hardened signals to military and government users only. Certainly, in the marine world, approaches such as trilateration of fixed AIS installations and/or RACONs (radar beacons) are being considered, but these techniques will not translate directly to land-based systems. Maybe we will see trilateration of cell towers used for GNSS backup, but that would leave huge areas with poor or no backup service.


There is a recent article about work towards a possible GNSS based Train Localisation On Board Unit (TLOBU), allowing ERTMS with many fewer balises, on the Inside GNSS website.

 

[najaB]

Not such an issue for Sat Nav systems, as they are medium earth orbiting satellites, where I understand there is plenty of room. The crowded part of space is for geostationary orbits, as that is a single line around the equator, so is most vulnerable to space junk.

Not really. In geo orbit the satellites are basically all going in the same direction at more or less the same speed so the likelihood of collision is low and the speeds involved will also be low. In MEO and LEO there are a mix of inclinations and orbits which means that closing velocities can be extreme. See the collision a few years back between an Iridium and Chinese satellite that generated thousands of new items of orbital debris.

 

[BayPaul]

Not really. In geo orbit the satellites are basically all going in the same direction at more or less the same speed so the likelihood of collision is low and the speeds involved will also be low. In MEO and LEO there are a mix of inclinations and orbits which means that closing velocities can be extreme. See the collision a few years back between an Iridium and Chinese satellite that generated thousands of new items of orbital debris.

It's beyond my area of expertise, but space is so big that surely the likelihood of collision is pretty small. Especially as there is ample redundancy for one or two satellites to be knocked out anyway. My fairly limited physics would say that high-speed items of debris are probably a lower risk, as they won't stay in the orbital plane for long, either flying into space or burning up in the atmosphere, whereas low speed debris in an orbital path will tend to hang around, with the potential for damage in a crowded bit of space.

A bit of looking on wikipedia suggests I am half-right. Geosynchronous is a major issue, but there are also some crowded Low earth orbit planes, where the high speeds cause a concern, especially, if I am reading between the lines correctly, directly over the poles, where many of the of planes cross.

 

[najaB]

My fairly limited physics would say that high-speed items of debris are probably a lower risk, as they won't stay in the orbital plane for long, either flying into space or burning up in the atmosphere, whereas low speed debris in an orbital path will tend to hang around, with the potential for damage in a crowded bit of space.

You might have misunderstood. All things in orbit are, by definition, high speed - around 7km/s in a typical orbit. What matters is relative velocity. An analogy: if two cars are travelling the same direction down the motorway both doing about 70mph and one drifts out of lane and they happen to touch there isn't likely to be much damage. Now contrast would would happen if two cars, both travelling at a lower speed (say 25mph) hit head on. The damage would be much more extreme.

The problem (fear!) with orbital collisions isn't just the damage done to the satellites that collide, but the fact that the debris from the collision will thrown out in all directions - which means further collisions, which results in further collisions. (See: Kessler syndrome).

A bit of looking on wikipedia suggests I am half-right. Geosynchronous is a major issue, but there are also some crowded Low earth orbit planes, where the high speeds cause a concern, especially, if I am reading between the lines correctly, directly over the poles, where many of the of planes cross.

That is why GEO is less of an issue - all the satellites in geo are in basically the same orbit. So, even if there is a collision the relative velocities are going to be much, much lower than they would be in LEO, where closing velocities can on the order of 15km/s. A 10g screw would deliver about the same amount of energy as half a kilogram of TNT.

 

[Irascible]

It's beyond my area of expertise, but space is so big that surely the likelihood of collision is pretty small. Especially as there is ample redundancy for one or two satellites to be knocked out anyway. My fairly limited physics would say that high-speed items of debris are probably a lower risk, as they won't stay in the orbital plane for long, either flying into space or burning up in the atmosphere, whereas low speed debris in an orbital path will tend to hang around, with the potential for damage in a crowded bit of space.

A bit of looking on wikipedia suggests I am half-right. Geosynchronous is a major issue, but there are also some crowded Low earth orbit planes, where the high speeds cause a concern, especially, if I am reading between the lines correctly, directly over the poles, where many of the of planes cross.

Collision in LEO is a huge concern, so much so that there's multiple organisations tracking every piece of debris down to as small as radar can pick up. Even a chip of paint at orbital speeds can be disastrous. There's a reckoning that it'd only take a couple more collisions ( or someone shooting a couple more satellites ) in LEO for there to be so much debris that we *couldn't go to space anymore*, because there'd be a chain reaction of debris hitting everything & causing more debris. Look up Kessler syndrome ( as above, now I read it ). The International Space Station regularily changes it's orbit to avoid debris, and that is an enormous structure by space standards.

Geostationary orbit is much much safer, not least because there's masses more of it - and of course if you park a satellite there it's going at the same speed as all the others ( or it wouldn't be stationary! ). They do sometimes oscillate vertically with respect to the earth's equator but laterally, no ( well not noticeably ). Once they're expired the sats are pushed out to a higher parking orbit - there's a possibility they might collide there, but the effect would be small anyway.

 

[D365]

As far as I am concerned, an ETCS is primarily a signalling system designed to provide the same or better protection than a conventional signalling system. Provision for some additional facilities has also been included.

Somewhat off-topic, but my understanding is that a primary benefit of ETCS Level 2 is that the braking curve of a train is tailored to the train, and the train’s speed, at any given moment. Rather than relying on fixed ground-based infrastructure such as TPWS.

 

[Annetts key]

Somewhat off-topic, but my understanding is that a primary benefit of ETCS Level 2 is that the braking curve of a train is tailored to the train, and the train’s speed, at any given moment. Rather than relying on fixed ground-based infrastructure such as TPWS.

Yep. ETCS includes ATP functionality. Once all trains that run on a particular line are fitted, you can then also remove the lineside signals, thus reducing the cost.
TPWS is an expensive cheap system with a lot of compromises.

 

[D365]

TPWS is an expensive cheap system with a lot of compromises.

I've heard some alternative (not forum-friendly) backronyms for TPWS

 

[edwin_m]

One of the GPS features being considered in the early days when I was involved was the "pseudo-satellite", a ground station that transmits "I'm a satellite , but my elevation is zero and my orbit is once per day"

== Doublepost prevention - post automatically merged: 3 Nov 2021 ==

Somewhat off-topic, but my understanding is that a primary benefit of ETCS Level 2 is that the braking curve of a train is tailored to the train, and the train’s speed, at any given moment. Rather than relying on fixed ground-based infrastructure such as TPWS.

It will also transmit the equivalent of signal aspects to the train almost instantly they change. So it avoids the situation where a driver passes a single yellow just before it was going to clear to a less restrictive aspect, and has to slow right down anticipating a red, until the next signal comes into view.

Yep. ETCS includes ATP functionality. Once all trains that run on a particular line are fitted, you can then also remove the lineside signals, thus reducing the cost.
TPWS is an expensive cheap system with a lot of compromises.

I was part of the group that developed the idea of TPWS, and played a small part in analysing all accidents back to about 1968 to work out how the sequence of events would have been different if such a system had existed at the time. This concluded that TPWS would have avoided about 70% of the casualties that a full ATP system would. This was treated somewhat cynically by Uff-Cullen, but subsequent events (or the lack of) have pretty much borne it out.

 

[Annetts key]

I’m not going off topic to fully debate ATP vs. TPWS again. But as far as the ground based equipment goes, the older ATP equipment is proving to be far more reliable than the newer TPWS equipment.