Re: [SystemSafety] Change of passenger warnings on Hamburg subway from today.

From: Matthew Squair < >
Date: Tue, 18 Feb 2014 16:22:06 +1100

Having travelled on the Moscow metro, where there's no doors closing alarms or announcements, I can tell you that Muscovites treat the car and station doors with all due respect, and so they should, no soft closing or obstruction detection on those babies.

So my prescription would be to remove the doors closing alarms and see what happens, both to your hit rate with passengers and timetable.

As I see it this is a risk homeostasis problem, making the situation appear to be riskier should therefore decrease the rate of passengers caught in the doors.

Matthew Squair

MIEAust, CPEng
Mob: +61 488770655
Email; Mattsquair_at_xxxxxx

On 18 Feb 2014, at 3:33 pm, Tracy White <tracyinoz_at_xxxxxx

I would tend to agree that there is likely to very little in terms of any variation in overall safety achieved. But I would be interested to know whether the time delay between the warning and door closure has altered in anyway as part of the change; it's a human behaviour thing, but if you can make it less attractive (rewarding) to try and jump in between closing doors, then speeding up the door closing process will allow less thinking time and less impetuous behaviour - you are making it harder for them to achieve the desired outcome and they will be less likely to exhibit that behaviour.

What I would add is that, from an engineering perspective, procuring a language/country independent component would be logistically attractive for both as a procurer and seller. Just wondering if we always have to make it about safety just to give it corporate emotional credibility.


On 18 Feb 2014, at 12:02, Steve Tockey <Steve.Tockey_at_xxxxxx

I doubt seriously this is going to make any difference. I frequent subway

systems around the world: Beijing, Shanghai, San Francisco's BART,

Washington DC Metro, Paris Metro, Tube in London, etc. In all cases,

whether the "doors closing" warning is verbal or beeps, it simply doesn't

matter. The frequency of someone getting stuck in a closing door is the

same everywhere. So while they may think it will make a difference, in

practice it won't. What they need to do is adjust their schedule to allow

for a certain percentage of such delays and just go with that. They will

have to deal with the delays no matter what...

-----Original Message-----

From: Karsten Loer <k.loer_at_xxxxxx

Date: Monday, February 17, 2014 12:58 AM

To: "systemsafety_at_xxxxxx


Subject: [SystemSafety] Change of passenger warnings on Hamburg subway

from today.

Dear colleagues,

after having enjoyed the inspiring discussions on this list for several

years now, mainly passively, I think I can contribute an interesting topic


Hochbahn, the operating company of Hamburg subway, decided that from today

on the natural language warnings "zurueckbleiben bitte" ("stay clear from

the train"), which used to warn passengers about closing doors of the

train that is imminent to depart, will be replaced by a series of beeps.

Warnings are broadcast at the end of the stop, and are initiated by the

driver after she/he has checked (by means of a set of monitors that are

installed in the cockpit and/or in an area of the platform where the front

of the train comes to a hold) that all doors are clear of passengers.

The old warnings, which had been in operation for several decades,

consisted of a natural language message that was broadcasted via the

public announcement system of the subway station. For newer train (series

DT4 and DT5) this was a recording in a friendly and polite female voice,

for the older DT3 train series these messages were read out by the train

driver at each stop (-- so as a side effect travellers were able to learn a

lot about the train driver's gender, nationality and current mood).

These warning messages were complemented by a series of short beeps at the

train doors.

The new warnings are just an extended series of short beeps at the train


The justification given by Hochbahn is as follows:

Apparently, over the years some people were conditioned to interpret the

old natural language warnings to stay clear of the train as a prompt to

speed up an try to get into the train.

As a consequence, people would get trapped in train doors; with a

potential of causing injuries to themselves and delays to the train (in

the new DT4 and DT5 train series the doors will open automatically, if a

certain degree of resistance is detected during the closing process; in

the old DT3 trains the doors need to be opened by the train driver, if

she/he becomes aware of the problem via video screen).

This causes a delay at the station. As a consequence, trains need to go

faster to catch up with the timetable. This, in turn yield an increase of

power consumption which accumulates to up to 700.000 EURO (about 1 Million

USD) per year. (Well, from personal experience asa frequent traveller on

these trains, I would have thought the main concern of faster going trains

would the safety of standing passengers who tend to bounce back an forth

due to the higher accelerations; but that matter is not mentioned anywhere

in the press statement.)

In summary, I see three arguments that are raised by the operator to

justify the need for the change of warning announcements:

(1) preventing passengers from getting trapped in doors of a train that is

about to leave,

(2) reducing the frequency of disturbances of the time schedule, as a

consequence of (1), hence, allowing trains on route to travel at lower

velocity and

(3) saving costs if operations as a consequence of (2).

As a safety analyst and concerned passenger I would like to add to this

list of arguments: (4) increasing the level of comfort/reducing the

potential of falls of standing passengers as a consequence of (2)

Perhaps there are other experts on this list who also doubt that changing

the warnings from a combination of spoken platform announcements and

audio-signals on the train to just audio-signals on the train will have

the desired effect?

I wonder how long it will take for people to adapt their habits and react

the door signal alone (Pavlov-effect)?

Maybe it is also worth thinking about the side conditions:

The subway trains in Hamburg operate very frequently and reliably; usually

every 3-5 mins during the peak times (7am through 9pm), and every 10 mins

at nighttime.

So one might wonder why people are prepared to take the risk of getting

injured despite the next train arriving soon.

I wonder if people's behavior would be likely to change if the imminent

arrival of the next train would be made visible, e.g. as is done on London

underground (although I am not sure if the motivation might be different).

In Hamburg this information is only displayed AFTER the current train has

left the platform.

Many greetings from Hamburg!

Dr. Karsten Loer

Dr. Loer system safety consulting

The System Safety Mailing List


The System Safety Mailing List


The System Safety Mailing List

The System Safety Mailing List
systemsafety_at_xxxxxx Received on Tue Feb 18 2014 - 06:22:19 CET

This archive was generated by hypermail 2.3.0 : Tue Jun 04 2019 - 21:17:06 CEST