Crowd Analysis (2017)

Next

Estimating crowd numbers, density and capacity is an essential requirement for crowd safety and risk analysis.

________________________________________________

“Everyone is entitled to his own opinion, but not to his own facts.” Daniel Patrick Moynihan

________________________________________________

Our analysis at Manchester Metropolitan University


Article in the New York Times

https://www.nytimes.com/interactive/2017/01/22/us/politics/womens-march-trump-crowd-estimates.html?smid=pl-share

________________________________________________

________________________________________________

Click here for the link on the Inaguration crowds 20th Jan 2017

As any crowd analyst will confirm, the front fills first. In the images below, both the 2009 inaguration and the 2017 inaguaration enents had capacity crowds near the stage (seated) area. 

From the rear view we can see that 2/3rds of the area in the 2017 inaguration event are empty.

________________________________________________

These two images (from youtube live feeds) show the difference of the two crowds attending the 2009 (Obama) and 2017 (Trump) inauguration. Marcel Altenburg and I were working with low resolution live feeds during the inaguration. Below are the analysis images we worked from during the event - live from 7 different video feeds to ensure these were both at the 45 minutes before the Oath of Office.




________________________________________________

Crowd Counting - in real-time

There are two types of event – planned and unplanned (organised and spontaneous gatherings). Planned events can be ticketed or free entry. Obviously, ticketed events allow an accurate capacity check – you calculate the safe holding capacity, number of seats/standing area and then sell that number of tickets.

So how do we calculate the crowd capacity and crowd numbers for the spontaneous and free entry events? We need to know the limits to avoid overcrowding and potential crown crushing. So how do we achieve this?

We start with the site area, the routes to/from the site and the safe capacity. That’s a simple calculation using various freely available tools – such as Google earth Pro – which allow the user to point and click an area, obtain a measurement of that area, and then to calculate a safe holding capacity. So, we can ascertain, before the event, how many people the site can safely accommodate.

Now, we need to count in real-time, to make sure we don’t exceed the safe limits. There are two methods of estimating crowd numbers – a count or a density check.

The first, is to count people in/out of an event space - this works well in closed spaces, such as a festival or mass gathering within a boundary, typically a fenced off area. Over the last decade, we have seen a variety of automatic counting systems, but none that provide a 100% accurate count at high volumes of throughput.

The second method, when you simply can’t count everyone (such as a network of streets/roads, open parks, and complex spaces or areas which have a high volume of input and output. These areas can be checked using a density evaluation - from a recent New York times Article - click here for a link. 

In realtime - we can use this to evaluate the crowd build up on the site - click this link for more information.

________________________________________________

New York Time article

From Lincoln to Obama, How Crowds at the Capitol Have Been Counted

People have descended on Washington to witness, celebrate and protest since the cornerstone was laid on the Capitol building in 1793. But estimating the size of those crowds is not easy, and sometimes crowd-counting can be fraught with controversy. Here’s how the tools for counting have changed over the last 150 years. An analysis by Dr. Keith Still and Marcel Altenburg estimated that 7,350 people are in this view:


We can estimate, using area and a packing analysis, coupled to an image head count to estimate crowd numbers. In the above example. between 7,050 (areas x packing density) and 7,350 (head count). Note: the size of people in 1861 was approximatley 20% smaller than 2017 human dimensions).

________________________________________________

Using real-time estimates for decsion support


Modelling the crowd build up for the Royal Wedding 29th April 2011.

Furhter information on this link

and for control room decision support - click here

________________________________________________

 Aritcle in the Washington Post (Jan 9th 2017)

Ignoring safety, what’s the maximum number of people that can fit in Times Square? Fewer than you might think. washingtonpost.com 

Click on the title above for the full article

________________________________________________

Below are two projects from our MSc student, Alexandar Kollaritsch, featured in the USA press.

http://www.post-telegraph.com/news/experts-broncos-parade-estimate-3-5-times-too-high/

http://www.post-telegraph.com/news/a-look-at-methodology/ 

http://www.post-telegraph.com/news/through-the-transportation-lens/


_________________________________________________________________

Manchester Met University Horizonal black logo

For further information on crowd density and crowd flow - click here.

We run training courses around the world and an APEL short course linked to the MSc in Crowd Safety and Risk Analysis at Manchester Metropolitan University in the UK.

Contact us for further information - click he

© G. Keith Still.  All rights reserved. All images, videos, audio files and text are protected under international copyright law  You may not copy, modify or use any of the website content, in whole or in part, for any commercial or non-commercial purpose without permission.  Contact email