What is Flow? The lean principle demystified

20images

20seconds

Any company

People & Resource

People & resource

People & resource

value value value

The basic principle of a company is that it increases the value of a product by applying people and

resources.

There are 2 ways to look at the efficiency of a company:

• We could look from the side, and observe if value is applied to the product all the time

• Or we could look from the bottom, and see if all people and resources are optimally used

Let’s look from the bottom – the resource efficiency view.

This machine is a Magneto resonance scanner – a very expensive machine.

Hospitals make it profitable by making sure it is being used all the time – causing waiting lists for the patients.

This is typical for resource efficiency: resources are busy all the time, but the flow object will move slower.

If we look from the side at the company we get the flow efficiency view.

In this case nobody is interested that this team of engineers is busy all the time.

What is most important is that the car can leave the pits as soon as possible.

This is typical for flow efficiency view: the flow object will move fast, but resources are not busy all

the time.

Assembly lines as Henry Ford introduced combine resource efficiency and flow efficiency:

The belt is moving all the time, and people are working all the time

The drawback of this system is that any kind of variance would stop the belt and efficiency gets very

low.

So let’s summarize:

• There are 2 ways to look at efficiency: flow efficiency and resource efficiency

• It is possible to achieve both at the same time, but in these cases your efficiency is very vulnerable

for variance.

Now that we have understood this, let’s look at the question ‘what makes things flow’?

Let’s first look at the flow object itself…

How can we shape it so that it flows easily?

These are marbles. They behave like a fluid since the particles are small, and have a low friction to a

surface. Also, they are all of similar size.

Can you see the similarities to user stories?c

So if we make our flow object fluid, do we get flow automatically?

Not quite. Something needs to make the flow object move.

Let’s call it potential energy.

In the picture here up to this point something has been pushing us – from now on it will be pull.

So is it up to us to decide if we apply push or pull as potential energy?

Sometimes yes.

In this case:no.

Overall you can say: whenever you can implement pull, you should do it.

Pull is easier when it comes to dimensioning bottlenecks. We’ll see this later on.

One thing is still missing: the flow object needs to flow THROUGH something – a medium.

A river is a suitable medium for the transport of water.

The river will make sure your water gets from A to B.

However, for different flow objects, you will need a different medium

Information could also be a flow object.

Rivers are not suitable for transporting information – this is a much more suitable medium.

The point is – you need to optimize your medium so that if fits your flow object.

Note that also information needs potential energy to flow: just ‘putting something on the web’ does

not make information flow.

Flow =

+ +

So let’s summarize what possibilities we have to make something flow:

• We can make the flow object ‘fluid’

• There needs to be potential energy: push or pull

• We have a suitable medium

Now that we have seen what makes things flow, let’s have a look on what can go wrong with flow?

Every process has got a bottle neck.

Maybe the bottleneck is not even a problem – if the throughput is low enough to fit through it.

This is also the advantage of a ‘pull system’ – the throughput cannot be higher than the bottleneck

allows.

But if the bottleneck IS a problem, what will happen to our flow objects?

What that can happen is what you see in this picture: queuing

Looking at resource efficiency, we would say the road is fully utilized so we are very efficient.

But of course in this situation, nobody is interested in resource efficiency.

But not only queuing can happen, this picture shows a different possible scenario: overflow

Depending on the nature of your flow object and medium, overflow occurs rather than queuing.

When overflow happens:

• The quality may be negatively affected since things get done too quickly.

• The medium gets damaged since e.g. people get stressed.

Is there anything else that can go wrong with flow, apart from obstructions or bottlenecks?

Here everything seemed to be in place:

• We have a fluid

• We have potential energy

• We have a good medium

But the temperature is not right.

In organizations , temperature means motivation, stress level, having the right tools etc.

But let’s assume that everything for flow is in place, including the right temperature.

Then, sometimes a situation like this happens: some flow object is not obeying the normal rules.

We often talk about the need for flexibility, but what we see here is that flexibility, at the same time, is

a variance that will potentially lower flow efficiency.

This is what happens if we only concentrate on flow optimization:

this may mean that some part of the system has at a certain time no flow units to process. This

effect is called ‘starvation’.

This does not mean that there is nothing to do. We can use this time to improve the flow of the

system.

In a flow optimized system, if one step in the system does not deliver in time, all steps behind them in

the flow would face starvation.

One way to avoid this is to introduce buffers in the stream

Buffers allow the downstream parts to continue to work for some time.

However, we need to be very aware that buffers increase the overall lead-time!

summary

H

i

n

d

e

r

e

r

s

So again let’s summarize what what hinders flow:

• Bottlenecks can cause queuing or overflow

• A wrong environment can completely stop the flow

• Exemptions and ‘flexibility’ impact flow directly

Finally, we have seen that we need to find the right balance between short leadtime and starvation

risk (flow vs. resource efficiency)

› Slide 1: Nearsblankpuzzle / Deviantart / CC-BY-SA-3.0

› Slide 3: Mj-bird / Wikimedia Commons / CC-BY-SA-3.0

› Slide 4: Bert van Dijk / Wikimedia Commons / CC-BY-SA-2.0

› Slide 5: Kyle Harris / Flickr / CC-BY-2.0 (cropped, recolored)

› Slide 7: caiophox / Deviant Art / CC-BY-3.0

› Slide 8: Jamie McCaffrey / Flickr / CC-BY-2.0

› Slide 9: David Precious / Flickr / CC-BY-2.0

› Slide 10: Signalhead / Wikimedia Commons / CC-BY-SA-3.0 (changed aspect ratio)

› Slide 11: Rafael Matsunaga / Flickr / CC-BY-SA-2.0 (cropped)

› Slide 13: werner22brigitte / pixabay / Public Domain CC0

› Slide 14: Christian Haugen / Flickr / CC-BY-2.0

› Slide 15: DVIDSHUB / Wikimedia Commons / CC-BY-2.0

› Slide 16: Frank Kovalchek / Flickr / CC-BY-2.0

› Slide 17: Thomas Leuthard / Flickr / CC-BY-2.0 (cropped)

› Slide 18: Josh Sommers / Flickr / CC-BY-2.0 (changed aspect ratio)

› Slide 19: Ikas.us / Wikimedia commons / CC-BY-2.0 (cropped)

› Slide 20: laobc / Openclipart / Public Domain CC0

attributions