How accurate is TRUST data?

[Peter Kazmierczak]

Just wondering how accurate the railway's internal TRUST data is?

[PaulRhB]

Very as long as the info is input correctly Like everything else it relies on humans too. No train should depart without its consist being uploaded as it's relied on for regulating trains. It's allowed us to spot out of gauge locos and stock and make sure trains fit in platforms or loops

If you find a train with no consist, especially freight with an unknown headcode, alarm bells ring. We always try to check specials and non WTT headcodes to check there's no wagons like loaded Warwells or mk3 coaches that have restrictions on parts of our patch.

[Dunsignalling]

Usually accurate, but not necessarily up-to-the minute as it shows where things were last reported. If that's moving vehicles in a train, that won't be exactly where they are right now and some rural sections involve 20 minutes or more of running time between reporting points. If even one automated reporting point goes down, things can easily "disappear" for the best part of an hour. Sometimes several in a row will go off-line for one reason or another.

 

Where arrivals/departures/passing times are not reported at a particular location in real time, the system usually won't catch up until the train arrives at the next functioning reporting point and/or somebody makes a manual input.

 

I know of more than one branch terminus which has no reporting equipment at all. Arrival times are input by the signaller at the controlling box, after the train terminates and the driver radios for permission to return. On the way back, an "on-time" from the first reporting point leads to an "on-time" departure being input for the end of the line, but some time after it actually occurred. In the absence of a report from the driver a late arrival will lead to a similarly late departure from the place of origin being input.

 

John

Edited April 27, 2018 by Dunsignalling

[Peter Kazmierczak]

Just asking, as I was enquiring on another site about some HST power car numbers and some of the info I received doesn't tally with what I actually saw.....

[PaulRhB]

That can be errors in input off the depot or a change of stock on the diagram later overwriting what you saw. If a train changes due to disruption or failure the new one will show on all parts of the diagram someone checking for you later sees different stock.

Edited April 26, 2018 by PaulRhB

[Peter Kazmierczak]

Thanks Paul.

That's probably it, as most of the info ties in with my sightings, but a few things don't.

[PaulRhB]

Control sometimes step up trains during an incident and don't tell Fleet straight away either so that can mean TRUST has several out of synch. Modern GPS tracking means fleet can sort it out once they know with some units or they rely on alerting staff to check numbers. It's important to sort it out fairly fast though or you can end up with diesels running low on fuel or stock over miles on service diagrams.

[Peter Kazmierczak]

Now I might've been given duff data, but suppose TRUST says 43133 + 43091 worked 1P30 when actually the power cars were 43133 + 43171. How would this be reconciled with engine hours / vehicle positioning?

[PaulRhB]

The depot would find out when the swap happened and calculate it from there working out where it had been. GPS tracking, where fitted, makes it a lot easier to work out later as it records it all plus allows remote monitoring of issues.

As most are hired from ROSCOs they are hired on mileage so they pay for any excesses of distance or missed service. It also means important maintenance periods are adhered to to prevent failures.

As only one unit is wrong it's possibly a depot input error as unusual to swap one power car without the depot updating it. Fleet will work it out though.

Mind you if there's gps tracking it would be easy to sort overnight so it may be only updated then.

Edited April 26, 2018 by PaulRhB

[Dunsignalling]

Now I might've been given duff data, but suppose TRUST says 43133 + 43091 worked 1P30 when actually the power cars were 43133 + 43171. How would this be reconciled with engine hours / vehicle positioning?

If stock gets substituted, especially en-route (i.e. a train getting terminated short and returning on a working other than that planned) data can often only be recalculated and manually input after the event.

 

Such alterations normally arise with very little prior notice as a result of "pressure" situations and control will usually have too much on their plate to be able to adjust diagrams quickly enough to allow the system to accept revised vehicle IDs from the commencement of the journey.

 

Most things, however changed from the plan, will get recorded in one or more systems, but the data won't necessarily migrate to TRUST of its own accord.

 

The main thing is, naturally, to get things moving asap and deal with any (virtual) clerical fallout after things quieten down. After a really big "nasty", some aspects might take a few days to get sorted, after any arguments over delay attribution have been settled, but the important stuff will be put straight within 24 hours.

 

John

Edited April 26, 2018 by Dunsignalling

[PaulRhB]

It will vary from TOC to TOC too what their recording method is for manual input. SWR fleet at Salisbury are very keen to keep it up to date due to outberthing on a regular basis meaning not all units get back every night.

[Banger Blue]

We use ‘Intergrale’ which superseded ‘Genius’ to keep track of the traction.

 

If we get a unit turn up on a working when we are expecting something else, it’s a quick call to the Train Service Controller, who will look at both units diagrams, find out where their paths would have crossed and the likely swap took place, then it’s just a case of reallocating the units from the appropriate point in the schedule and the system does the rest adjusting the fuel & exam miles accordingly.

[Alien]

Just wondering how accurate the railway's internal TRUST data is?

 

In regard to delays and TDA its not accurate at all in "some" cases because, due to the present day system, profit comes before everything else...........

[Rivercider]

I thought that, strictly speaking, TRUST only covers the running times and delays,

a separate system shows what stock is actually on the train, albeit linked to TRUST.

When I was still working in Bristol TOPS, (up until 1996) it was POIS that was used to input details of HST sets and diagrams,

as well as loco hauled sets.

 

Also running times will not get reported on TRUST if a train gets diverted off the booked route,

 

cheers   

[PaulRhB]

Yes various systems feed into TRUST but it's the only one widely available to Signalboxes, station staff, Control etc, I guess Peter has a source with access to that.

It's that part that's not meant to be made public which is why you don't see freight headcodes on the maps of realtime trains either as it all comes under business confidentiality and you have to be a paid up FOC or TOC to access it so outsiders don't know exactly what formations they are running. Obviously it also has security implications with certain trains and people have got into a lot of trouble posting that!

Genius is still in use and something called Exalla, though the spelling may be wrong , and others as mentioned above. I guess TOCs each have their preferred modern systems for miles and servicing.

If it's not on realtime trains I don't post it, not worth risking the job over it

Edited April 27, 2018 by PaulRhB

[caradoc]

In regard to delays and TDA its not accurate at all in "some" cases because, due to the present day system, profit comes before everything else...........

 

Can you clarify that statement please ? Are you really saying that the data entered into Trust by Attribution Staff is falsified, if so that is a sweeping and possibly slanderous accusation; Mistakes can and do occur, but Train Operator Controls (TOC) have real-time access to Trust and can, and do, dispute any attribution they believe to be incorrect. Both Network Rail and TOC Controls have replay access to CCF (showing train movements and in some cases signal aspects) and make full use of this facility. If the matter cannot be resolved at Control level it then goes to the next level, the Performance Sections.

 

The present day system at least ensures that the running of every single train is accurately recorded and reasons for any delay of 3 minutes and over provided; That was not the case in the 'good old days' of BR, for example before Trust the arrival time of Down WCML express trains at Glasgow Central was extrapolated from the passing time at Cambuslang and 7 minutes added; If a delay between Cambuslang and Central occurred, unless Control was specifically advised, it went unrecorded and unexplained.

Edited April 27, 2018 by caradoc

[Rivercider]

I can image that more TRUST delays might be apportioned to whichever TOC/FOC/NR fights the corner less skillfully.

 

I remember in the early days of TRUST when almost all TRUST locations were manual input.

I worked in Bristol TOPS, our 'passenger' desk was linked in to the Temple Meads radio net and we did all the 

inputs of times, and allocated delay codes as we saw fit, or as directed by the platform staff, or occasionally the Duty Station Manager.

 

I can remember, particularly in times of disruption how many possible reasons for delay there might be. 

A theoretical example a Gloucester to Taunton stopping train losing time at Temple Meads.

It is late from Gloucester, due to late inward working, which could be for any number of reasons.

There are a few more minutes of delay en route due to TCF or signal problems before arrival at Bristol.

The forward driver is delayed into Bristol working up from the west, for other reasons.

Due to congestion there is a platform change at Bristol, so passengers have to use the subway, and off other late trains, so there is now more delay.

Now the train is held to follow a XC service for the west.   What code to use? 

I would input the code I thought most appropriate, and sometimes checked later to see if control had amended it.

 

cheers

[w124bob]

This an actually conversation at New st on Saturday just before the christmas break about 6 years ago, I'd arrived at New st about 40mins early and made several attempts to ring control eventually getting through.

Control "Bob we think there's a train on 9 for Picc, can you take it back"

Yes there was a full voyager on 9 and it had been there sometime.

Second call from control " whats the unit number?"

When doing set swaps where 4 morphed into 8 or 9 cars I would always ring the the box manager at New st or Piccadilly to give him the new train lenght, it was always appreciated as XC were not always good at updating TRUST during a last minute set swap/change. As a driver I just did it to minimise hassle arriving on occupied platforms at journeys end at Picc.

Edited April 27, 2018 by w124bob

[APOLLO]

Page 12 explains TRUST

 

https://cdn.networkrail.co.uk/wp-content/uploads/2016/11/Catalogue-of-Railway-Code-Systems.pdf

 

Brit15

[Alien]

Can you clarify that statement please ? Are you really saying that the data entered into Trust by Attribution Staff is falsified, if so that is a sweeping and possibly slanderous accusation; Mistakes can and do occur, but Train Operator Controls (TOC) have real-time access to Trust and can, and do, dispute any attribution they believe to be incorrect. Both Network Rail and TOC Controls have replay access to CCF (showing train movements and in some cases signal aspects) and make full use of this facility. If the matter cannot be resolved at Control level it then goes to the next level, the Performance Sections.

 

The present day system at least ensures that the running of every single train is accurately recorded and reasons for any delay of 3 minutes and over provided; That was not the case in the 'good old days' of BR, for example before Trust the arrival time of Down WCML express trains at Glasgow Central was extrapolated from the passing time at Cambuslang and 7 minutes added; If a delay between Cambuslang and Central occurred, unless Control was specifically advised, it went unrecorded and unexplained.

 

Dear Caradick,

Your reply sounds like you are a Manager or a Controller. Given the accusations you have made against me, clearly you believe that someone is guilty based on assumption and there is no need to await any evidence. I hope you don't get called for Jury service anytime soon.

 

I have simply replied to a question from the OP with my opinion. It wasn't directed at you, so you should desist from making such allegations against me in this way. 

 

I have previously enjoyed this using forum but it would appear it is not the amiable place I thought it was.

[PaulRhB]

Guys If there are issues with attribution there are methods to report and deal with it rather than arguing here and in justifying it could breach social media policies.

[Dunsignalling]

I haven't tried it recently, simply because I retired, but entering a 'V' enquiry into the TRJA screen used to bring up loco/unit, coach and sometimes wagon numbers/length/weight. As I worked a passing loop on a single line, and the few freight moves we did get were diversions, length was always the most important piece of information!

 

Mileage and maintenance info wasn't included.

 

John  

Edited April 28, 2018 by Dunsignalling

[PaulRhB]

Yes that's still correct John, TRUST is very much a Signalmans friendly tool for planning possible regulating with both times and whether it will fit! When John was on the WoE there was a very good regulating telegraph with all the boxes talking and working together to get the best result. Panels have the advantage of a big overview but single boxes could work trains just as efficiently when they talked. CCF and TRUST as a bonus give all the boxes the view so they have advance warning that makes the overall running miles away visible. What may seem odd locally can make good crosses further away that recover the service much faster. With the right decisions on turnarounds we always reckoned we could recover the service in 1-2 hours once we got rid of the problem.

Length plays a huge part in not locking up a station where it's just a loop or more complex.

In the Down direction we only have 1000ft to play with to loop a freight as there are only the main and platform loop available. In the Up direction we can use the Reception and Up yard and have over twice that so can loop even the high output track trains.

[caradoc]

Dear Caradick,

Your reply sounds like you are a Manager or a Controller. Given the accusations you have made against me, clearly you believe that someone is guilty based on assumption and there is no need to await any evidence. I hope you don't get called for Jury service anytime soon.

 

I have simply replied to a question from the OP with my opinion. It wasn't directed at you, so you should desist from making such allegations against me in this way. 

 

I have previously enjoyed this using forum but it would appear it is not the amiable place I thought it was.

 

Your extremely witty and clever misspelling of my user name shows the level of debate you wish to engage in, perhaps when complaining that the Forum is not the amiable place you thought it was you should look in the mirror ?

 

And yes, I have many years experience as a Controller, whose duties including providing information to Attribution staff, at times inputting to Trust Attribution myself, and dealing with disputes and complaints regarding attribution. Your OP, provided without any supporting evidence whatsoever, was a direct attack on those engaged in the Attribution process.

 

However, I have no wish to see you abandon this Forum and look forward to the evidence you will provide to support your assertion that

'In regard to delays and TDA its not accurate at all in "some" cases because, due to the present day system, profit comes before everything else...........'

[beast66606]

FOC do not always update loco locations in a timely manner - DRS and GBRF being two that can be a bit slow at updating.