Notting Hill Gate incident

The RAIB report into the Clapham South trap-and-drag incident in March 2015 was published in March 2016. It contained the following recommendation:

(italics in original)

The status of this recommendation is marked as "Implemented" in the June 2017 Index of RAIB recommendations (report 04/2016 on page 63), but there is no additional detail provided in the linked Reconsideration(s) and Status Report.

Given that this evidently did not prevent the incident at Notting Hill Gate, I would not be surprised if the RAIB take another look at LU's PTI management.

Frankly there's nothing that can be done to prevent accidents like this that isn't going to cost vast amounts of money that TfL simply does not have

Very true.

My personal view is this is one area where people need to take responsibility for themselves. No matter how observant drivers are and no matter how good the safety systems, the fact remains that these incidents will happen very occasionally. People can minimise their risk by not boarding or alighting a train when the doors are closing, instead of blindly running for a train door even when there’s another train a minute or two behind. A culture change is necessary on this.

I don’t think a guard is necessarily a solution. There’s probably more overall risk having a guard hanging out the door of a moving train, or taking up a vestibule and having to direct people away from using the local door (thus placing them in a conflict situation with no immediate place of safety available), than the current setup. The most we would get is a DLR style setup where the local door is closed and then the bell given. Perhaps the guard might get a set of CCTV monitors, but in reality this is no different to a Central Line driver using ATO. It doesn’t do anything to address that things are generally busier than in the past and people seem to more prone to injudicious actions nowadays, not to mention very much more likely to be distracted by devices.

In the case of the DLR incident, the passenger wasn’t doing exactly what they should have been doing if they were running.

One could remove a lot of dragging risk if people knew that most LU trains have doors which can relatively easily be pushed open a few inches for the very reason of being able to remove objects. Trouble is if it became common knowledge then it would be misused. Sadly at times thentravelling public are their own worst enemy and everything has to be designed around the basis people are dumb animals.

Every weekday morning I stand near idiots who have limbs, bags, clothing etc well over the yellow line, inches from the side of a train whipping in or out at speed. I'm surprised there aren't far more incidents like this.

To be clear, my post was not advocating a reintroduction of guards just the introduction of some method of alerting a driver that there is a need to stop the train due to a problem on a platform they cannot see.
Surely it would be technologically possible for a CSA to be given a button they could press to send an emergency stop message to the driver of a train departing that platform via the radio.

You just need a device that has some method of sending a single message over the radio - it can either send to all trains in the station area (no more disruptive than any other broad emergency stop message) or, if it knows which platform it's on (see below for three possible ways of doing this) then it can send the message to train in a specific location.
1 - Manually entering the platform's identity into the device (using the three letter code for the station and up to three characters for the platform you can uniquely identify every platform in the country, not just on LU - if that's too complicated then just two characters for the platform and tie the device to a specific station).
2 - Wifi - the identity of the base station connected to can be used to identify which platform its on. You might not be able to distinguish between different faces of an island platform, but it requies the least input from the CSA.
3 - The technology used to confirm security guards have patrolled an area (I don't know the name) - just tap the device against the metal stud to sign on to the platform.
If the radio technology would be too bulky for hand-held then the device could just a message to a box on the platform (which could even do the knowing where it is bit) which sent the radio message, adding milliseconds of delay.

Stop plungers are a bit of an anachronism now. LU has decided they are not worth the cost and complexity, not to mention the risk of misuse. One wonders if the ones which still exist may quietly get decommissioned at some point.

The sensitive edge equipment on the S stock seems to achieve this, all be it in a very different way from described.

Indeed, but it will be many, many, years before all trains are so equipped. I haven't heard of any plans to retro-fit such equipment to the current - non S stock - fleet. (Which doesn't mean it's not being considered, of course, but I imagine it would be difficult in a number of ways.)

Of course, fitting cameras above a few hundred platforms and linking them up wouldn't be a walk on the park.

I doubt whether the CSAs doing SATS can see more than the driver during the peaks, they are at platform level with passengers packed along the platform while the drivers get a view over the heads of the passengers from at least four different cameras.  In the case of Notting Hill Gate its six and I'm not sure the CSA can even see the entire platform from any one spot.

Connect radio is quite simple and I doubt if it could support complex additions but when we have emergencies there's the "Mayday" button, helpfully coloured red for quick identification.  The line controller has a queue of calls waiting on their screen, when someone presses a "Mayday" button that call goes to the top of the queue in big red letters.  If for example there was an obstruction on the Chiltern Line which runs parallel to the Central Line West Ruislip branch we would have to contact Wood Lane control room who would then pass the information onto Network Rail unless there is a NR signal telephone handy. 

The only way train drivers can talk to station staff is to scroll through the channels on Connect until you find whichever station you want to talk to which is an extremely lengthy process, its actually quicker to call Wood Lane to get them to contact the station staff, get out on the platform to use a help point or walk to the station control room to find someone to talk to.  The only way a CSA would be able to communicate with a train would be to phone them having identified the leading cab number 

As North End said LU has decided emergency stop plungers are not worth the cost and complexity, if they don't consider it worth spending money on an existing system then they aren't going to spend on a new system

if they don't work they should be removed or at least clearly marked as out of use. We don't want people trying to use them in an emergency instead of something that actually works.

In my mind, the potential harm from these types of incidents occurs in seconds.
Combining this with human factors (even CSAs are human, and a person getting dragged by a train is a rare & startling occurrence, but something getting stuck in a door is a common occurrence - it would take somebody stopping the train within seconds to avert serious injury.
Even the time taken to reach an emergency stop plunger could be too long. The only use I could see for the plunger is stopping train from entering a platform if someone is on the track, or preventing a train from moving if someone falls under it while it is stationary. Stopping a train that has started to move would take too many seconds to avoid injury to the person, run to the nearest plunger, fiddle about with the flap, press the button - the train might be almost out of the station by then.

The people best placed to act might be other passengers, pulling a PEA.
My understanding is that only one door leaf in a double set can be pulled back, so there is a 50/50 chance of a trapped object pulling against the door that can be pushed back depending on which direction the train is travelling in.

Whatever the system would have to be simple and automatic. Until trains all have sensitive edges, maybe there isn't much that can be done.
Apart from increasing vigilance and training for staff and drivers (e.g. to observe unusual movements of passengers).