The Railway Children

Recently, I had the opportunity to have a look around a railway network control room. It looked a bit like this:

railway control roomPretty impressive eh? What struck me was what a fantastic example of a system the railway network is. Think about it – all those trains having to be coordinated so as to ensure the overall system functions effectively. Train crews need to be transported around the network in time to operate other train routes; certain trains need to be in certain places at certain times to facilitate the effective running of these routes. Every part of the system depends on every other part.

On top of that, different services sometimes need to have their carriages coupled and uncoupled at particular stages of their journey to enable trains to be split and sent in different directions. The arrival of services at railway station platforms must also be coordinated, taking into account the size of available platforms and many other factors. Signalling and other equipment has to operate effectively and trains must pass through points at the right moment to prevent congestion or delays.

If an incident occurs that causes a delay, then it’s almost inevitable that there will be a knock-on effect. The challenge then is to minimise the potentially cumulative impact of such an incident. This could involve re-routing particular services, or slowing others down so that they arrive at different points at different times and in a different order to the original plan. Essentially, the result is a dynamic reconfiguration of the system’s activity, intended to minimise disruption following an unforeseen event.

This is such a great example of a system because it highlights the importance of interdependencies, with the screens in the control room visually displaying real time performance information. If one small part of the system fails, there is likely to be an impact somewhere else in the system. If a state of balance is not restored quickly, then further consequences occur as other parts of the system begin to fall out of synch. The result is an accelerating snowball effect that can cause absolute chaos (as many travellers will be aware).

Having that ‘whole system’ view is necessary to maintain an optimal balance, i.e. homeostasis. However, in many traditional organisational structures, this crucial perspective can be obscured due to silo-based models and fragmented working practices (see this blog post for more), resulting in sub-optimisation and a breakdown in cohesion. Worse than that, where counterproductive performance management practices (such as the use of arbitrary numerical targets and league tables), and superficial use of data (e.g. binary comparisons) are present, this inadvertently encourages behaviour that damages the system and destabilises the natural balance required for it to function effectively.

As a result of not understanding systemic interdependencies and clumsy use of performance data, the risk of fundamentally misguided interference by managers is heightened, leading to knee-jerk reactions and an untold cycle of damage.

Staying with the railway theme, someone recently likened this behaviour to letting a bunch of children loose in an old-style signal box, pulling on the levers randomly without understanding what the levers do – perhaps even thinking they are making a positive difference. Meanwhile, the wrong points are being opened and closed, trains are being sent in the wrong direction, and maybe some are even crashing into each other, far away down the line where the consequences cannot be seen from the signal box.

railway signal levers

The question is this:

“Which approach to managing the system is dominant in your organisation?”

About InspGuilfoyle

I am a serving Police Inspector and systems thinker. I am passionate about doing the right thing in policing. I dislike numerical targets and unnecessary bureaucracy.
This entry was posted in Systems thinking and tagged , , , . Bookmark the permalink.

8 Responses to The Railway Children

  1. neil says:

    The railways are designed so that even if children were allowed to run amok, there would be little chance of an accident.

  2. Tim says:


    Good blog, but though I get that this is a analogy, works as such, it’s not fully representative.

    Though there are a number of variables within a rail network system, it is more ‘closed’ than say a social based system. I liken it in simplistic terms to a continuum on which say a person’s weight is the most ‘closed’ ie you pretty much have full control over the variables & feedback loops (measures in your world) are fast & easy to decipher – has weight reduced/increased?

    Though not simplitistic by any means, the rail network is (in my view more towards the closed than most would think), it is the reason why the Japenese have such good control over theirs.

    Social based systems by their nature are just about as open as can be, multifaceted, loose connections, long feedback loops. It’s very difficult to find interdependencies & when you do they often change or you are actually looking at symptoms not causes. But I like the anology, in re to that people need to take a systemic view, not a reductionist view.

    It’s why I don’t understand people going over board re waiting time targets (they are wrong by the way), it’s not the big picture. We are living longer – tick, but how is how we are living causing strain on critical services, how do we then intervene in the right place (clue, it’s not at the hospital level).

    Think you would enjoy Ashby’s work – he’s spot on in re to design, it’s a theoretical framework but gives the underlying thought process to drive design work.

    Keep up the blogs, I enjoy them

    • Yep, you’re totally right. It would be a great start though if some managers even (erroneously) thought about their complex systems as closed ones rather than not as a system at all…

  3. Richard B says:

    I can’t just march into the boss’s office and tell him his doing it all wrong!
    So what are the correct measures to assess policing?

    • Why not?

      Correct measures in policing depend on what you’re trying to find out about – e.g. crime rates (if accurately recorded) are useful as a source of info for deploying resources, but not as a measure of police ‘performance’. End-to-end times for investigations, arrival times at emergencies and proportion of incidents resolved at the first point of contact are all helpful measures.

  4. Mark Patel says:

    Hallelujah, I wish folks in the NHS would realise this is exactly how hospitals function.

    I use the metaphor of the human body (any biological system will do, and clinicians relate better to the human body then other systems). They intrinsically know when happens if you run a biological system at 100% or close to maximum. They also know that there are key indicators in the human body that will tell you if something if a person is healthy or sick.

    And when you get a signal a person is sick, then further diagnostics are needed to identify the illness, before a treatment can be applied. Once treatment is applied you need to check back against those indicators to see if its working. If not, you change the treatment, learning from what you’ve already tested.

  5. Change Officer says:

    Great post, although I do remember seeing a poster proudly displayed at my local station saying how the train company had met its target of 95% of trains arriving on time. Why would anyone aspire to be late 5% of the time?!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s