Class 701 'Aventra' trains for South Western Railway: progress updates

[Goldfish62]

What’s Freudian slip?

An unintentional error that reveals subconscious feelings.

 

[WWTownEnth]

Anyone know which unit has just emerged from Long Marston?

 

[800001]

Anyone know which unit has just emerged from Long Marston?

Surprisingly 701520/522 Long Marston to WD

 

[Peter Sarf]

Surprisingly 701520/522 Long Marston to WD

Gulp, its happening, scraping further from the bottom of the other barrel !.

It was going to be 701045 last week so I wonder if that refused too much ?.
Or did SWR not make enough room for it (10car vs 5car sidings) ?.

EDIT
To add 701520 and 701522 only left Wimbledon for Long Marston on 06/08/2025 so might well be healthier than the last two 701/0s at Long Marston.
701045 arrived December 2024 from I know not where.
701049 arrived January 2025 from I know not where.

 

[800001]

Gulp, its happening, scraping further from the bottom of the other barrel !.

It was going to be 701045 last week so I wonder if that refused too much ?.
Or did SWR not make enough room for it (10car vs 5car sidings) ?.

EDIT
To add 701520 and 701522 left Wimbledon for Long Marston on 06/08/2025 so might well be healthier.

The only 9 car left at LM is 701049.

 

[mansonlo49]

The only 9 car left at LM is 701049.

isn’t 701045 still at Long Marston?

 

[Peter Sarf]

isn’t 701045 still at Long Marston?

Yes refused last week.
Or maybe that is what the other 57 is for !!!!.
except RTT shows 5Q70 as double headed.

 

[TT-ONR-NRN]

Gulp, its happening, scraping further from the bottom of the other barrel !.

How do you mean - isn’t it a good thing?

 

[Peter Sarf]

How do you mean - isn’t it a good thing?

Not a bad thing in itself except it feeds my fear that the remaining 701/0s are getting harder and harder to get into traffic so progress for the whole fleet of 90 units is looking harder. We have seen progress on 701/0s stutter of late (possibly Christmas).

I feel there maybe be plans afoot for 701/5s soon !.

 

[Goldfish62]

Not a bad thing in itself except it feeds my fear that the remaining 701/0s are getting harder and harder to get into traffic so progress for the whole fleet of 90 units is looking harder. We have seen progress on 701/0s stutter of late (possibly Christmas).

I feel there maybe be plans afoot for 701/5s soon !.

There have been two 701/0s enter service this month, although 701038 failed on its first outing and hasn't been seen since.

While in normal circumstances it would be sensible to get all of one sub-class commisioned first I agree that there may come a time, possibly soon, that in order to make any sort of progress the 701/5s start being put into service.

 

[800001]

isn’t 701045 still at Long Marston?

Hadn't realised it had stayed there.

 

[Ian Hardy]

The only 9 car left at LM is 701049.

Don't you mean 10 car, as I didn't know that SWR were expecting any 9 car units.

 

[800001]

Don't you mean 10 car, as I didn't know that SWR were expecting any 9 car units.

Yes, I meant 10 car, a typo, simple mistake!

 

[TEW]

There have been two 701/0s enter service this month, although 701038 failed on its first outing and hasn't been seen since.

While in normal circumstances it would be sensible to get all of one sub-class commisioned first I agree that there may come a time, possibly soon, that in order to make any sort of progress the 701/5s start being put into service.

The plan still remains Summer for 701/5 introduction after the 701/0s are all introduced. Some extra 701/5s are required to start the necessary testing prior to their introduction though, which I believe is why this pair are making their way South.

 

[cduncan]

No 455s allocated in the updated CWN. 1N91 1713 Waterloo-Frimley/Aldershot should be allocated 10 701 but so far has been 8 450 this week. Only 1N90 0619 Farnham-Waterloo remains 8 450.

 

[Peter Sarf]

After the delivery today (Wednesday 11/02/2026) using 57303 & 57306 RTT only shows 57306 going to Wembley. Wonder what 57303 is up to next ?.

Realtime Trains | 0M70 2150 Wimbledon Park Depot Sdgs to Wembley Loco Sidings P,q,r | 11/02/2026

Real-time train running information for 0M70 2150 departure from Wimbledon Park Depot Sdgs to Wembley Loco Sidings P,q,r on 11/02/2026. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

Below are the preceding moves.

To save anyone looking the locos are 57303 & 57306. I was hoping to find one was a 99 or better still both were - but no.

RTT link below.

Realtime Trains | 0V70 0851 Leicester L.I.P. to Long Marston Gbrf | 11/02/2026

Real-time train running information for 0V70 0851 departure from Leicester L.I.P. to Long Marston Gbrf on 11/02/2026. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

Nothing to say both locos are for 701 moves, nothing to say where to, nothing to say one loco is chaperoning the other loco after a major repair/fault, could be two locos needed to end up in SWR land so one is hitching a ride.

It has been known for one 701 to leave for Wimbledon and one 701 to leave for Eastleigh. But I think that was one loco for both jobs and both from nearby Marchwood.

Looking at RTT there is only the one actual move today and the timing load is same as last weeks cancelled one.

Realtime Trains | 5Q70 1435 Long Marston Gbrf to Wimbledon Park Depot Sdgs | 11/02/2026

Real-time train running information for 5Q70 1435 departure from Long Marston Gbrf to Wimbledon Park Depot Sdgs on 11/02/2026. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

 

[wickham]

After the delivery today (Wednesday 11/02/2026) using 57303 & 57306 RTT only shows 57306 going to Wembley. Wonder what 57303 is up to next ?.

Realtime Trains | 0M70 2150 Wimbledon Park Depot Sdgs to Wembley Loco Sidings P,q,r | 11/02/2026

Real-time train running information for 0M70 2150 departure from Wimbledon Park Depot Sdgs to Wembley Loco Sidings P,q,r on 11/02/2026. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

Below are the preceding moves.

Did both locos work to Wimbledon - did ayone observe the train ? It is reported elsewhere that 57303 was detached at Worcester - presumably for another job.

 

[ls2270]

Did both locos work to Wimbledon - did ayone observe the train ? It is reported elsewhere that 57303 was detached at Worcester - presumably for another job.

57303 does indeed show on TOPS as at Worcester Yard.

 

[Peter Sarf]

57303 does indeed show on TOPS as at Worcester Yard.

Thanks.

 

[norbitonflyer]

[Fredian slip] An unintentional error that reveals subconscious feelings.

In this case, (Corodia) associating Coradia with corrosion

 

[NSEWonderer]

Not been a good day for SWR after 2pm due to a incident at clapham on platform 8, points failure at windsor, a 458 having a bit of a smokey situation and then 701031 failing at earlsfield on the down: https://www.realtimetrains.co.uk/service/gb-nr:P62109/2026-02-12/detailed which is now 5M61, detrained and will be heading outt he way after being stuck for several minutes

 

[mansonlo49]

There’s a drag next Monday to Wimbledon, haven’t seen one on Monday before as it’s usually on Wednesday

Realtime Trains | 0V70 0851 Leicester L.I.P. to Long Marston Gbrf | 16/02/2026

Real-time train running information for 0V70 0851 departure from Leicester L.I.P. to Long Marston Gbrf on 16/02/2026. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

Realtime Trains | 5Q70 1435 Long Marston Gbrf to Wimbledon Park Depot Sdgs | 16/02/2026

Real-time train running information for 5Q70 1435 departure from Long Marston Gbrf to Wimbledon Park Depot Sdgs on 16/02/2026. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

 

[Nimbus020]

Not been a good day for SWR after 2pm due to a incident at clapham on platform 8, points failure at windsor, a 458 having a bit of a smokey situation and then 701031 failing at earlsfield on the down: https://www.realtimetrains.co.uk/service/gb-nr:P62109/2026-02-12/detailed which is now 5M61, detrained and will be heading outt he way after being stuck for several minutes

Thanks for this - what was the incident at Clapham Plat 8? - I resorted to buses home last night as it looked chaos towards Wimbledon in the evening

 

[norbitonflyer]

Thanks for this - what was the incident at Clapham Plat 8? - I resorted to buses home last night as it looked chaos towards Wimbledon in the evening

Delays around that time are variously shown on RTT as due to trespass and a person hit by a train. Probably the same incident as one can often lead to the other.

 

[NSEWonderer]

Thanks for this - what was the incident at Clapham Plat 8? - I resorted to buses home last night as it looked chaos towards Wimbledon in the evening

One under as norbiton mentioned, right around evening peak, followed by several other incidents.

 

[w11_supremecy]

Thanks for this - what was the incident at Clapham Plat 8? - I resorted to buses home last night as it looked chaos towards Wimbledon in the evening

A person jumped in front of 2L46 just before 15:00 yesterday. Caused absolute mayhem for the rest of the day, and it wasn’t helpful that as NSWWonderer said, a train broke down on the Down Main Slow, and a 458/4 had a fire of some sort at Wandsworth Town. Bit of a mad one.

 

[swtrains]

It is very very likely the station names on the 701s are exactly what the customer asked for. I can imagine this was quite possibly an existing list got from somewhere by the customer, possibly a list of what Desiros already had on them, possibly an old list of what was originally wanted on trains, possibly an abbreviated list knocked up quickly and never padded out correctly by the customer or possibly not !. I can imagine the customer got exactly what they asked for in the fianl specification before programming work started.

Now lets not forget the list of station names might be realistically viewed as not much more than data so potentially not hard to fix for the 701s. But I can imagine the software might not cope with the services that loop back to Waterloo and this might require quite a big change to the logic in the software making it very different to that used on other Aventras (as in costs money).

Its easy to blame the result but not so easy to consider that the request was poor.

Are you suggesting the shortened station names are what was asked for, when they're not present on all the 701s that have not been updated?

On the original production release for revenue service back in January 2024, the only shortened name was Lon W'loo, and that was only visible on the internal PIS when showing the calling pattern. This was updated fairly quickly afterwards to what most units are running now, with no shortened names. So why would they deliberately reintroduce them?

As I posted earlier in this thread, the shortened names were shown at Derby in the testing lab. It was confusing at the time, but I expected the actual update to use the full station names, alongside an improved PIS. This was not the case. This is the exact same version that was shown in August 2025, with seemingly no visual changes since. Happy to post more photos if needed.

As we know, the 701s have the worst PIS software out of all the Aventras, although it's confusing as to why they could not copy some basic screens from one version to another. The 730s, for example, have the correct TfL lines (Elizabeth line, not "TfL Rail") and TfL font styling. Adapting that for SWR should be trivial - swap the service status text to SWR's font - but there's clearly no incentive to go beyond the brief. You'd expect these things to be sorted by now.

For context on why the PIS is so poor: the software that runs on all the Aventra PIS displays is NOT written natively. It's essentially a Chromium web instance running on the screens - infamous for being a memory hog - which explains the sluggishness and the blank screens for several seconds at each station while it loads a webpage. Compare that to the far superior 707s, which ran natively written software in C/Qt, on Linux. The Aventra displays are likely running full Windows (as seen on a 345 once), which, on underpowered hardware, just compounds to the lag.

On top of that, the software isn't even developed in the UK - it's outsourced to Alstom in India, meaning developers cannot physically test on the trains. The 701 PIS system reportedly went through over 1000+ revisions just to reach the January 2024 state. And from press release footage, Derby doesn't even use real PIS destinations on the test track - it shows generic placeholders like "Station 17 via Station 14." How that constitutes adequate testing is beyond me.

The reality is the PIS will never be truly good until it's rewritten natively, which almost certainly won't happen given the outsourcing. They can refine what exists, but when a year-long update cycle produces subtle fixes (coach 6-10 text no longer cut off, Clapham interchange lines now centred) alongside regressions (shortened station names, triple-repeated announcements, screen freezing, incorrect information), the outlook isn't great - especially when Alstom's testers apparently don't catch any of it.

As for looping services, it would honestly be simpler to do what the 455s and 450s do: the guard sets a PIS code for Kingston loopers showing Strawberry Hill/Teddington as the destination, then enters a new code at STW/TED to switch it to Waterloo. Trying to handle that in software would probably take them years and break something else in the process.

 

[Gag Halfrunt]

On the original production release for revenue service back in January 2024, the only shortened name was Lon W'loo, and that was only visible on the internal PIS when showing the calling pattern. This was updated fairly quickly afterwards to what most units are running now, with no shortened names. So why would they deliberately reintroduce them?

If a dummy dataset is used for testing, perhaps someone installed the wrong dataset (the one with shortened names) on the latest version after it had been tested.

Then again, perhaps the PIS software is so badly designed that station names are hardcoded instead of being read from a separate file and any change to the names requires a software update and another test cycle.

 

[Peter Sarf]

Are you suggesting the shortened station names are what was asked for, when they're not present on all the 701s that have not been updated?

I overlooked that. In which case I smell bad version control (see below). To be fair maybe a compromise.

On the original production release for revenue service back in January 2024, the only shortened name was Lon W'loo, and that was only visible on the internal PIS when showing the calling pattern. This was updated fairly quickly afterwards to what most units are running now, with no shortened names. So why would they deliberately reintroduce them?

As I posted earlier in this thread, the shortened names were shown at Derby in the testing lab. It was confusing at the time, but I expected the actual update to use the full station names, alongside an improved PIS. This was not the case. This is the exact same version that was shown in August 2025, with seemingly no visual changes since. Happy to post more photos if needed.

As we know, the 701s have the worst PIS software out of all the Aventras, although it's confusing as to why they could not copy some basic screens from one version to another. The 730s, for example, have the correct TfL lines (Elizabeth line, not "TfL Rail") and TfL font styling. Adapting that for SWR should be trivial - swap the service status text to SWR's font - but there's clearly no incentive to go beyond the brief. You'd expect these things to be sorted by now.

I fully expect there are now different versions of the software for each class of Aventra. So no longer possible to roll out the same changes for all classes. I used to spend time trying to get back to one program for two or more departments. It was quite risky and hard when comparing differing programming styles/approaches.

For context on why the PIS is so poor: the software that runs on all the Aventra PIS displays is NOT written natively. It's essentially a Chromium web instance running on the screens - infamous for being a memory hog - which explains the sluggishness and the blank screens for several seconds at each station while it loads a webpage. Compare that to the far superior 707s, which ran natively written software in C/Qt, on Linux. The Aventra displays are likely running full Windows (as seen on a 345 once), which, on underpowered hardware, just compounds to the lag.

Oh dear. probably generates a huge piece of software that covers all eventualities most of which are impossible.

Yes, its possible a new version has hit a limit in terms of size and so names might have had to be shortened to buy some space. I even remember the transition from 3GL to 4GL in the 1980s created bloated software !.

On top of that, the software isn't even developed in the UK - it's outsourced to Alstom in India, meaning developers cannot physically test on the trains. The 701 PIS system reportedly went through over 1000+ revisions just to reach the January 2024 state. And from press release footage, Derby doesn't even use real PIS destinations on the test track - it shows generic placeholders like "Station 17 via Station 14." How that constitutes adequate testing is beyond me.

So the developers are working at a distance that precludes local understanding and easy nuanced communication. Twas ever thus.

Conjecture BUT I have seen a regression or two in my time. Often because quick easy and obvious changes were made and released while someone-else / different-team was making a more involved, important and so slow, changes. Both working from the SAME previous version. When the more involved change finally went live the other changes were of course missing !. The more involved change might have taken priority so not necessarily a mistake, maybe a conscious decision to re-do the easy changes again later (or sometimes not bother in my experience).

The reality is the PIS will never be truly good until it's rewritten natively, which almost certainly won't happen given the outsourcing. They can refine what exists, but when a year-long update cycle produces subtle fixes (coach 6-10 text no longer cut off, Clapham interchange lines now centred) alongside regressions (shortened station names, triple-repeated announcements, screen freezing, incorrect information), the outlook isn't great - especially when Alstom's testers apparently don't catch any of it.

I agree. Do it properly instead of using a lazy development tool that generates bloated software. Visual Basic .net versus C springs to mind. You can tell I am old.

Occurs to me that "subtle fixes (coach 6-10 text no longer cut off," means the software was first written for the five car units only. Although of course the ten car units are actually two five car units permanently (outside a depot) coupled together. So coaches 6 to 10 are in the "other" unit.

The parent unit (leading and/or with the driver in) might be passing coach names to the slave unit with extra info pertaining to the unit address - perhaps this stole some of the available name space ?. Again just a guess based on my experiences in the past sorting out software/system bugs when I was counted on as an overzealous perfectionist !.

As for looping services, it would honestly be simpler to do what the 455s and 450s do: the guard sets a PIS code for Kingston loopers showing Strawberry Hill/Teddington as the destination, then enters a new code at STW/TED to switch it to Waterloo. Trying to handle that in software would probably take them years and break something else in the process.

I agree. Sometimes better to let humans do it rather than get too clever with software.

== Doublepost prevention - post automatically merged: 13 Feb 2026 ==

If a dummy dataset is used for testing, perhaps someone installed the wrong dataset (the one with shortened names) on the latest version after it had been tested.

Then again, perhaps the PIS software is so badly designed that station names are hardcoded instead of being read from a separate file and any change to the names requires a software update and another test cycle.

Would not surprise me.

 

[norbitonflyer]

As for looping services, it would honestly be simpler to do what the 455s and 450s do: the guard sets a PIS code for Kingston loopers showing Strawberry Hill/Teddington as the destination, then enters a new code at STW/TED to switch it to Waterloo. Trying to handle that in software would probably take them years and break something else in the process.

The screens on the platform already say Richmond before they reach Wimbledon, and say Waterloo (via Richmond) before they reach Kingston. Simiklarly in the other direction, at Richmond they are shown as being for Wimbledon. Announcements on the train vary.