Fatal Accident at Ickenham station (28/03/2025)

[ainsworth74]

I couldn't immediately find a thread relating to this incident but the RAIB have now published a report into it today:

Summary​

At around 22:30 on 28 March 2025, a passenger fell from the platform onto the track at Ickenham London Underground station. Unable to get to a position of safety, the passenger lay on the track undiscovered for around 2 minutes before being struck by an arriving train, which then stopped normally in the platform.

Following the stop, the train’s brakes automatically applied as it left the station due to part of a safety system on the train coming into contact with the passenger. Subsequent investigations into the cause of this brake application led to the discovery of the passenger, around 14 metres from where they had fallen onto the track. The passenger was fatally injured.

RAIB’s investigation concluded that the passenger lost their balance on the platform, causing them to fall onto the track. Staff on duty at the station were unaware that the passenger had entered the station in a vulnerable state, or that the passenger was on the track after they fell. This meant that there was no intervention that prevented trains entering and departing from the platform.

CCTV evidence suggests that the passenger was aware that a train was approaching and that they were probably attempting to move towards the platform face and out of the path of the train. However, the under-platform recess was occupied by communication cables and this meant that there was insufficient space to accommodate the passenger and to allow a train to pass without contacting them.

The train operator also did not see the passenger on the track. This may have been because they were focused on stopping their train at the correct location and monitoring the platform‑train interface. The contrast between the brightly lit platform and dark track bed and the passenger’s dark clothing may also have impacted the train operator’s ability to see the passenger.

RAIB identified two underlying factors. The first, a probable underlying factor, is that London Underground’s standards relating to under-platform recesses were not being complied with and were not consistent with each other. The second, a possible underlying factor, was that London Underground had not completed platform-specific risk assessments for most platforms on its network or identified the safety benefit of some measures intended to mitigate the risk of people falling from platforms and subsequently being struck by trains.

Recommendations​

RAIB has made three recommendations, all addressed to London Underground. The first recommends that London Underground review its standards relating to under‑platform recesses to ensure that they are effective and consistent. The second recommends site-specific risk assessments for all station platforms and the implementation of appropriate risk controls. The final recommendation relates to providing operational staff with the necessary guidance and training to safeguard people under the influence of alcohol on the London Underground network.

Two learning points have been identified. The first is how a thorough track check can be effective in preventing further train movements when a person has fallen onto the track, as was the case in this accident. The second highlights the importance of following company policy with regards to routine drug and alcohol testing following serious accidents.

Report 02/2026: Fatal accident at Ickenham London Underground station

RAIB has today released its report into a fatal accident at Ickenham London Underground station, 28 March 2025.

www.gov.uk

I feel like from a brief skim there's some similarities here to that equally horrible incident at Stratford where a passenger fell onto the track and wasn't discovered until after they'd been run over by a train.

 

[Royston Vasey]

I feel like from a brief skim there's some similarities here to that equally horrible incident at Stratford where a passenger fell onto the track and wasn't discovered until after they'd been run over by a train.

I thought I was reading a post about that incident, thinking "wasn't that Stratford?" until I read your sentnce above. It's remarkably similar.

 

[deltic]

The bizarre point about this unfortunate incident is that the victim is reported as having a glass of wine, walked from work to the station in a reasonable time and then started acting in a manner as if they were severely intoxicated with the medical report stating that a large amount of alcohol had been consumed shortly beforehand. It would suggest that either they had consumed significantly more than a glass of wine, their drink had been spiked or what was thought to be wine was actually a spirit of some type. Something clearly doesnt stack up.

 

[Haywain]

It would suggest that either they had consumed significantly more than a glass of wine, their drink had been spiked or what was thought to be wine was actually a spirit of some type. Something clearly doesnt stack up.

I'm no expert but I would have thought that if their drink had been spiked it wouldn't show up as a large amount of alcohol having been consumed.

 

[AlterEgo]

The bizarre point about this unfortunate incident is that the victim is reported as having a glass of wine, walked from work to the station in a reasonable time and then started acting in a manner as if they were severely intoxicated with the medical report stating that a large amount of alcohol had been consumed shortly beforehand. It would suggest that either they had consumed significantly more than a glass of wine, their drink had been spiked or what was thought to be wine was actually a spirit of some type. Something clearly doesnt stack up.

The report says the toxicology report indicates they were absolutely battered from having been on a bender, basically. The anonymous witness evidence can be discarded as partial and unreliable. Paras 68-70.

 

[Benjwri]

I'm no expert but I would have thought that if their drink had been spiked it wouldn't show up as a large amount of alcohol having been consumed.

Extra alcohol being added to a drink is the most common form of spiking. That being said it’s extremely unlikely someone would be able to give someone enough alcohol to cause a blood alcohol that high without them noticing.

 

[61653 HTAFC]

I'm no expert but I would have thought that if their drink had been spiked it wouldn't show up as a large amount of alcohol having been consumed.

Depends what it was "spiked" with. I've always taken "spiked" to mean that something was nefariously added- this could be some sort of illicit sedative, or as simple as a spirit such as vodka. Obviously the former wouldn't show up in alcohol screening but the latter would.

Another possibility with a sudden display of intoxication disparate with the amount of alcohol that witnesses saw consumed, is low blood sugar. This can lead to very erratic behaviour and a loss of co-ordination not dissimilar to alcohol intoxication. Excessive alcohol can bring the blood sugar levels down to dangerous levels, so a combination of those two factors is a possibility.

 

[Benjwri]

Another possibility with a sudden display of intoxication disparate with the amount of alcohol that witnesses saw consumed, is low blood sugar. This can lead to very erratic behaviour and a loss of co-ordination not dissimilar to alcohol intoxication. Excessive alcohol can bring the blood sugar levels down to dangerous levels, so a combination of those two factors is a possibility.

Low blood sugar is known to cause false positives on a breathalyser, but it cannot change the blood alcohol content itself and will not affect testing described in the report. The level of BAC mentioned in the report unfortunately points pretty conclusively to more alcohol having been consumed than reported by the witness.

 

[61653 HTAFC]

Low blood sugar is known to cause false positives on a breathalyser, but it cannot change the blood alcohol content itself and will not affect testing described in the report. The level of BAC mentioned in the report unfortunately points pretty conclusively to more alcohol having been consumed than reported by the witness.

I'm not questioning the blood alcohol level, nor breathalyser results- I'm talking about behaviour here. It's clear from the toxicology that the unfortunate individual had consumed more alcohol than the witnesses believed. Someone had suggested spiking, but low blood sugar combined with very severe intoxication could have contributed to the individual ending up on the tracks. Alcohol also lowers the blood sugar (particularly if the type of alcohol consumed is low in sugar, e.g. vodka) and low blood sugar can lead to an increase in erratic behaviour.
Without wishing to speculate too much, someone who imbibes as much alcohol as this person did is probably not at "his first rodeo", but this time he ended up on the tracks. It's possible that a number of biological factors were at play.

 

[deltic]

The strange point is that they could walk from work to the station in a reasonable time but then once at the station they were all over the place and took time to get over the stairs and to the platform. Suggests they consumed a large quantity of alcohol, knowingly or unknowingly in a very short space of time as the toxicology report indicates which is at odds with the witness statement that they consumed a glass of wine between 22:00-22:16. Do pubs still allow their staff to drink on the job these days?

 

[aar0]

The strange point is that they could walk from work to the station in a reasonable time but then once at the station they were all over the place and took time to get over the stairs and to the platform. Suggests they consumed a large quantity of alcohol, knowingly or unknowingly in a very short space of time as the toxicology report indicates which is at odds with the witness statement that they consumed a glass of wine between 22:00-22:16. Do pubs still allow their staff to drink on the job these days?

That’s what stands out as odd to me. Some pubs do, but they don’t allow up to get drunk.

The under platform refuges getting blocked up by cables and also not being required on new platforms is interesting. Seems like a zero cost thing to include in design and build, and low cost in terms of avoiding putting cables there.

 

[waterboo]

What stuck out for me was once again LULs failure to comply with post incident for cause testing on the Train Operator

 

[DavyCrocket]

What stuck out for me was once again LULs failure to comply with post incident for cause testing on the Train Operator

And zero action other taken from after the Waterloo death

 

[boiledbeans2]

[...]

I feel like from a brief skim there's some similarities here to that equally horrible incident at Stratford where a passenger fell onto the track and wasn't discovered until after they'd been run over by a train.

Also, a similar incident where a woman was ran over twice on the Northern Line at High Barnet:

Woman hit by two Tube trains takes TfL to High Court

Sarah de Lagarde lost an arm and leg after being struck by two Tube trains at High Barnet station.

www.bbc.co.uk

 

[Mawkie]

Also, a similar incident where a woman was ran over twice on the Northern Line at High Barnet:

Not really, with Ickenham, Waterloo, and Stratford the victims were well over the alcohol limit for driving - and in the case of Hillingdon, so drunk it could have been life threateningly so for some.

Sara wasn't drunk, and TfL withdrew their suggestion that she may have been.

It's very hard as a member of staff to deal with severely intoxicated people - but it's harder to deal with the aftermath of a fatality.

Having said that, I have been in situations where I've been alone on a station and have challenged an apparently intoxicated person. They are often aggressive and abusive when told they can't travel, frequently threatening violence and being intimidating. So I do understand why they go unchallenged by gateline staff or lone working Supervisors. There simply isn't the support around these days from other staff, or BTP if things get nasty. It's a horrible situation to feel so isolated whilst running a station.

In my own case I have previously escorted intoxicated passengers to the platform and waited with them for a train, found out where they're going and arranged for them to be met at their destination. It's not the proper thing to do, but it's the best I could do at the time.

After Stratford (where you'll remember the victim was asleep on the platform for some time, stood up and became disorientated and fell on the tracks) as a driver I now call in to the line controller to report any sleeping passengers on platforms. I presume (and hope) they call the station and send some staff down.

I often hear "TfL didn't do this or that', but every incident is throughly investigated internally and often externally too. TfL have a" Vison Zero" policy towards injury (some call it Zero Vision!) but is that just corporate speak and do we have to accept some level of injury and death when moving the volume of people that TfL move?

I certainly go to work with the safety of all passengers and staff as my top priority, and indeed it drives every decision I make, but what is the actual practical solution to people falling onto tracks and getting struck by a train?

 

[waterboo]

Not really, with Ickenham, Waterloo, and Stratford the victims were well over the alcohol limit for driving - and in the case of Hillingdon, so drunk it could have been life threateningly so for some.

Sara wasn't drunk, and TfL withdrew their suggestion that she may have been.

It's very hard as a member of staff to deal with severely intoxicated people - but it's harder to deal with the aftermath of a fatality.

Having said that, I have been in situations where I've been alone on a station and have challenged an apparently intoxicated person. They are often aggressive and abusive when told they can't travel, frequently threatening violence and being intimidating. So I do understand why they go unchallenged by gateline staff or lone working Supervisors. There simply isn't the support around these days from other staff, or BTP if things get nasty. It's a horrible situation to feel so isolated whilst running a station.

In my own case I have previously escorted intoxicated passengers to the platform and waited with them for a train, found out where they're going and arranged for them to be met at their destination. It's not the proper thing to do, but it's the best I could do at the time.

After Stratford (where you'll remember the victim was asleep on the platform for some time, stood up and became disorientated and fell on the tracks) as a driver I now call in to the line controller to report any sleeping passengers on platforms. I presume (and hope) they call the station and send some staff down.

I often hear "TfL didn't do this or that', but every incident is throughly investigated internally and often externally too. TfL have a" Vison Zero" policy towards injury (some call it Zero Vision!) but is that just corporate speak and do we have to accept some level of injury and death when moving the volume of people that TfL move?

I certainly go to work with the safety of all passengers and staff as my top priority, and indeed it drives every decision I make, but what is the actual practical solution to people falling onto tracks and getting struck by a train?

I do feel for the drivers involved in such situations. The nature of stopping at stations (fast approaches, small margins of error in stopping locations, and PTI monitoring), makes it difficult to divulge attention to the trackbed during station approaches.

Doesn't the DLR have some sort of trackbed obstacle detection system that brings the train to a stand automatically? I know on Light Rail Blackpool trams was actively testing a pedestrian collision avoidance system. Could such technologies be retrofitted onto LUL?

 

[Pegpilot]

Waterboo, I know that LUL were looking at obstacle detection in stations about a decade ago, but that was in the context of thinking about driverless trains. I helped them translate a German technical standard describing just such a system, using multiple IR beams crossing the 4-foot as I recall, cleverly designed to be triggered by an object at least as large as a football to avoid nuisance triggers. Clearly the detection of an object would apply the EB on a CBTC system, but how you would do it with legacy fixed lineside signals ? Dunno.

 

[Mojo]

This FoI response from just a few days ago states that TfL have been trialling for the past few months “smart-detection technology” to identify people on the track at West Ruislip, East Acton, and Mile End. With a different system already in place on the DLR network at Custom House and further trials due at Pudding Mill Lane and Camden Rd (Overground).

== Doublepost prevention - post automatically merged: 6 Mar 2026 ==

Clearly the detection of an object would apply the EB on a CBTC system, but how you would do it with legacy fixed lineside signals ? Dunno.

It would alert staff in a control room who could then manually radio trains to get them to stop (or approach at caution, as appropriate).

 

[leonie13]

Sara wasn't drunk, and TfL withdrew their suggestion that she may have been.

I was of the understanding TfL withdrew this suggestion because they couldn't prove the individual was intoxicated - not because the individual definitely wasn't. The individual survived and therefore no toxicology report was administered/required.

 

[Haywain]

This FoI response from just a few days ago states that TfL have been trialling for the past few months “smart-detection technology” to identify people on the track at West Ruislip, East Acton, and Mile End

This would potentially be an application for CCTV based AI monitoring of platforms.

 

[CarrotPie]

This would potentially be an application for CCTV based AI monitoring of platforms.

Oh no, not more AI surveillance...

 

[35B]

I was of the understanding TfL withdrew this suggestion because they couldn't prove the individual was intoxicated - not because the individual definitely wasn't. The individual survived and therefore no toxicology report was administered/required.

If this is the case I'm thinking of, TfL were also very vulnerable to accusations of trying to smear the individual in question.

 

[Haywain]

Oh no, not more AI surveillance...

You'd prefer not to save lives?

 

[AlterEgo]

Oh no, not more AI surveillance...

This technology would survey predictable and static images like trackbeds and alert someone when the image changed. It may well save a few lives a year and is one of the most ethical uses of AI I can think of.

 

[bramling]

I do feel for the drivers involved in such situations. The nature of stopping at stations (fast approaches, small margins of error in stopping locations, and PTI monitoring), makes it difficult to divulge attention to the trackbed during station approaches.

I get somewhat fed up with these investigations. Trains are generally not driven on sight, and I feel it’s wrong for the driver to be on the receiving end of investigation when the track should be clear. If this had been on a station where trains non-stop then it’s all rather moot anyway.

Furthermore, in the majority of these cases the root cause is that someone has intoxicated themselves. Whilst the High Barnet one is supposedly contentious, no one has provided a credible alternative.

 

[AlterEgo]

I do feel for the drivers involved in such situations. The nature of stopping at stations (fast approaches, small margins of error in stopping locations, and PTI monitoring), makes it difficult to divulge attention to the trackbed during station approaches.

This incident also took place in the dark and the deceased was wearing dark clothing. Driver didn't really have a hope of seeing them.

 

[Watershed]

I get somewhat fed up with these investigations. Trains are generally not driven on sight, and I feel it’s wrong for the driver to be on the receiving end of investigation when the track should be clear. If this had been on a station where trains non-stop then it’s all rather moot anyway.

Furthermore, in the majority of these cases the root cause is that someone has intoxicated themselves. Whilst the High Barnet one is supposedly contentious, no one has provided a credible alternative.

The investigation didn't assign any blame whatsoever on the driver. But it did find that TfL hadn't followed its own standards, and had in fact recently added non-compliant cabling under the platform. This may well have been the last "hole in the Swiss cheese" that resulted in the passenger losing their life.

I'd say that's very strong justification for looking into this, even if it may have seemed a straightforward case of "passenger was drunk" at first glance. Moreover LU hardly comes out of this smelling fresh as roses when this is the third time their failure to perform D&A test has featured in a RAIB report in the last few years. There's no suggestion the driver was impaired in this case, but it's a routine procedure for a reason.

Moreover, anyone can fall off a platform; it doesn't just happen to drunk people. It could be the result of a disability, a malicious act or overcrowding. The report RAIB just published on the foreseeable crowding at Ealing Broadway caused by the poor design of the newly installed waiting shelter is instructive on that point.

Given the substantial risk that those falling onto the tracks face, the least that can be done is to adhere to existing standards when making changes, to reduce the chance of them suffering life-changing injuries or worse.

 

[35B]

The investigation didn't assign any blame whatsoever on the driver. But it did find that TfL hadn't followed its own standards, and had in fact recently added non-compliant cabling under the platform. This may well have been the last "hole in the Swiss cheese" that resulted in the passenger losing their life.

I'd say that's very strong justification for looking into this, even if it may have seemed a straightforward case of "passenger was drunk" at first glance. Moreover LU hardly comes out of this smelling fresh as roses when this is the third time their failure to perform D&A test has featured in a RAIB report in the last few years. There's no suggestion the driver was impaired in this case, but it's a routine procedure for a reason.

Moreover, anyone can fall off a platform; it doesn't just happen to drunk people. It could be the result of a disability, a malicious act or overcrowding. The report RAIB just published on the foreseeable crowding at Ealing Broadway caused by the poor design of the newly installed waiting shelter is instructive on that point.

Given the substantial risk that those falling onto the tracks face, the least that can be done is to adhere to existing standards when making changes, to reduce the chance of them suffering life-changing injuries or worse.

Not to mention that the consistency of those standards was lacking, with different regimes in two different documents. And let’s not comment on saying that space isn’t necessary because an unrealistically small 6 year old wouldn’t fit in the left over space after cables were installed.

 

[JohnofWessex]

Looking at not only the reports mentioned, a recent inquest into somebody drowned in Oxford, deaths or missing people in the Bristol Avon etc etc it seems to me that there is a major issue around deaths associated with intoxication.

Thats before we start on Drink Driving, committing and being a victim of crime etc.

While it is outside the RAIB's remit I suggest that taking steps firstly to try and reduce intoxication and a robust approach towards those intoxicated in public (Drunk Tanks anyone) might save lives

 

[Haywain]

While it is outside the RAIB's remit I suggest that taking steps firstly to try and reduce intoxication and a robust approach towards those intoxicated in public (Drunk Tanks anyone) might save lives

Asking as someone who enjoys a few beers, how do you propose to do that?