by Romain Bievelez

Image from www.brannlaw.com

A human crowd is very complex and fascinating. It can shape the world during elections (for the best or the worst). It can save lives and it can kill. Information and diseases can propagate through it and it behaves very specifically during waves of panic. Scientist have been now studying human crowds for decades. It may surprise you but it turns out that physicists are the foremost type of scientists that study human crowds, more than psychologists and biologists!

In this blog, you will learn about all sort of characteristics and behaviours of crowds. You will understand why crowds are studied mostly by physicists and you will learn about the impact that crowd behaviour can have.

But first, before jumping into this amazing topic, let us first define what is actually meant by “crowd”. It doesn’t have to be physical or tangible. A group of people chatting on social networks is here considered to be “a crowd”.
A human crowd is a group of people that are somehow interacting with each other.

Now that we know what a crowd is, we can dive into learning how it behaves in some classic situations.

Physicists started by studying movements in crowded areas. It has been seen very quickly that crowds move like fluids or granular solids (like sand or gravel). They literally “flow”. It comes from the fact that, in some sense, humans repel each other in the street. Obviously, when you are walking on the pathway, you don’t want to bump into the person that is coming in front of you. So, when the collision seems imminent enough, you deviate from your path to the easiest alternative way. This very simple behaviour (which is shared by a huge amount of social animal species) leads the crowd to flow like a fluid.

Image of a crowd behaving like a fluid from www.gae-engineering.com

This “fluid physics” description can be very accurate and lead to useful conclusions. For example, did you know that, in a crowd during a panic wave, you must avoid at all cost the edges of the crowd? You must follow the flow and stay in the middle of “the wave” to avoid getting injured!
Why is that? It is because the pressure gets bigger on the edges of a fluid. In other words, the shocks become more frequent and violent near the walls of the container. These could be the walls of the room and the buildings in the street.

Here is another fact: did you know that putting an obstacle in front of an emergency exit actually speeds up the evacuation? Yes! It is counter-intuitive. But like sand flowing through a hole. If you put a thin object in front of the door, it actually “breaks the flow” and help avoid congestion at the door.

Another similarity between fluids and crowds is that some things can propagate and diffuse through them. For example, fake news and diseases can propagate through a human crowd and it is a certainty that describing and understanding their spreading will help predicting and preventing popular misconceptions and pandemics.

But how can we do that? How can one predict how a disease will spread into the population? Actually, we haven’t reached this capacity yet. But we know how to reach it. We need to “model” disease spreading as accurately as possible. Modelling means basically to describe something using math. Let’s say you want to model the quantity of virus among the population. What you want is actually a function of time giving the quantity of virus. In other words, you want a mathematical rule that gives the quantity of virus if you give it a date.

Image of a crowd simulation software from www.lsi.upc.edu

Nowadays, scientists can predict much more than just the number of virus. To do so, they use advanced mathematical tools such as the so-called “discrete diffusion equation in finite space”. This long name refers to a mathematical rule that describes how anything can propagate (or diffuse) into a space that as boundaries (like a room for example). With such tool, we are now capable of accurately predicting how a virus would spread among a small group in a room!

Crowd study is what we call an “empirical” science. It means that experiments play a central role in the development of this science. But what does it mean to “measure a crowd”? The most common ways of measuring a crowd are, first, simply giving a survey to all the participants of the experiment and, second, recording the crowd on camera and then analysing the recordings. Most of the time, physicists use this second option. It is perfectly suited for the study of movements as we have now advanced software that can track people on an image. It than also allows to measure the so-called “crowd density”: the number of people per square meter which is so important to assess injury risks in a crowd.

But then, how can we measure large population movements like migrations for instance? Physicists have now discovered that seismology could be used as a non-intrusive tool for population tracking.
Basically, we can now identify the human influence in seismic noises recorded on ground vibration captors all around the world. This will be used to record population migration over thousands of kilometres without actually breaking into people’s private life.

Crowds can actually become useful. For example, it has been discovered that a crowd of doctors is very effective at avoiding antibiotic over-prescription.
Studying crowds has also applications in other sciences that are related to population study. A lot of tools developed for human crowd studying are used now to study other animals. We can now track the movements of flocks of bird, herds of sheep and shoals of fish. We can dynamically control flocks of drones and take effective measures against the spreading of diseases.

It makes no doubt that crowd study is a fascinating science with loads of applications and utilities and it has a great future with plenty of discoveries.

References:

“The physics of fluids explains how crowds of marathon runners move” by Emily Conover, ScienceNews:

https://www.sciencenews.org/article/physics-fluids-explains-how-crowds-marathon-runners-move

“Math to match pedestrian behavior is all about timing” by Andrew Grant, ScienceNews:

https://www.sciencenews.org/article/math-match-pedestrian-behavior-all-about-timing

“Solution to century-old math problem could predict transmission of infectious diseases”, ScienceDaily:

https://www.sciencedaily.com/releases/2020/05/200529150617.htm

“Harnessing the ‘wisdom of crowds’ can help combat antibiotic over-prescription”, ScienceDaily:

https://www.sciencedaily.com/releases/2020/11/201103115656.htm