The science of walking in crowds

Pedestrians scurry across Queen Street in the heavy rain. Photo / Jason Oxenham

What can the way we move in crowds tell us about how we might design better cities? Nickolas Morton is studying physics at the University of Auckland, and worked at Te Punaha Matatini, a national research centre that studies complex systems and networks, over the summer. He talks to science reporter Jamie Morton about a new paper he's just published.

Q. You've just published an interesting study looking at the link between binary decision-making and breaking symmetry. Can you walk me through it?

A. Imagine you are walking down a busy street, Queen St for instance.

When confronted with someone moving in the opposite direction, you have two choices - you either pass on the right, or you pass on the left.

This is a binary decision.

But the other pedestrian has a similar choice: if you both choose the same, you'll pass each other easily.

If you make opposite choices, you might walk into each other.

Over time you might develop your own preference for passing other pedestrians by learning what other people around you prefer.

Perhaps you'll learn to favour the left more than the right: in physics we would say this breaks the symmetry between left and right.

Over time, this might result in a collective norm being established, so that everyone who walks down Queen St will come to prefer the left over the right, or vice versa.

This seems to be what happens, but different communities develop different preferences - the English prefer to pass on the right while Kiwis tend to pass on the left.

Our study looks at what might happen when groups of pedestrians meet but have distinct preferences: say All Blacks and England fans gathering at Twickenham before a rugby test.

We are interested in whether this new group of pedestrians will arrive at a new collective consensus: whose preference would win out, or would chaos result?

Obviously if one group outnumbers the other, then this group's preference is likely to become established.

But we also found that if the crowds are more evenly balanced and they ignore the preferences of others, then the flow of the crowd can be impeded as ordered flows of pedestrians break down.

Q. When you're trying to pick apart these patterns, how complex does the math get?

A. Symmetry breaking isn't a new concept, it's been around for decades.

We've borrowed the maths to describe this from the physics of magnets - magnets have a north pole and a south pole that arises from a collective preference of the electrons.

The harder part is describing the patterns of the pedestrians themselves.

In our case we model the pedestrians as point charges being repelled by a magnetic force - our model people don't like other people in their personal space.

Again, the mathematics is well understood, which makes it easy to solve using a computer.

Q. To physicists, why is pedestrian movement such a great case study?

A. Physics is all about understanding the universe around us.

What's interesting, is that when we apply equations we've learnt from our understanding about physical, material systems, like magnets, to macro biological and social systems, the equations still often work.

In essence, the behaviour of atoms can be used to describe the behaviour of humans.

People are more complicated than atoms though: people try to anticipate what other people will do.

This is important in our model.

If our pedestrians are able to guess correctly what others will do, they can avoid collisions and move more easily through crowds.

Q. How does this picture vary in different communities?

A. When communities are isolated from each other the symmetry breaking that occurs in one community can be different to what happens in another.

For instance, in Japan, citizens of Osaka have a different side of the escalator that they like to stand on than citizens from Tokyo.

This causes some angst when people from Tokyo meet people from Osaka in train stations.

There are many more examples, most of which are country to country differences.

Q. In a "real world" sense, how might these understandings help urban design? And what happens when planners go against the flow?

A. The easiest example would be a hallway or small room inside a crowded art gallery or museum - somewhere where you would expect mixing between tourists and locals who might have different preferences.

If planners ignored this mixing between the two norms, congestion could occur before a new collective norm was established.

In order to counter that they could increase the size of the hallway or room to reduce the effect.

Or they could simply try to bias pedestrians decisions in the form of a sign, such as "keep left."

Some museums seem to have already caught on to the congestion crowds can induce, and have solved it by making their entire museum full of loops, so that the crowd always flows in the same direction.

However, the Wellington city public library has its entrance on the right, which often catches locals out, because it is encouraging a crowd flow that runs against the local preference.

Q. Aside from pedestrians, what other examples can or have been used to study the very same principles?

A. One of the other researchers at Te Punaha Matatini, Dion O'Neale, has used a similar model to look at how fashion changes.

Some people want to dress the same way as everyone else so that they fit in, while the trend setters want to stand out from the crowd.

Any situation like this, where your best choice depends on what you think others will do, can exhibit similar behaviour.