Why crowd counting and test events are so important

Measuring crowds accurately is quite a challenge, especially at events. In the media, you often hear about different visitor numbers for the same event, depending on who is being interviewed: organiser, mayor, security or safety services. In a way, this is not very surprising. After all, you can count people in different ways, both manually and with the help of technology. But which method tells the truth? Time for research! We follow the researchers during their measurements at the first test conference in our country and at the weekly market in Bornem.


In cooperation with IDLab-imec of the University of Antwerp, the centre of expertise Public Impact of KdG University of Applied Sciences and Arts studies the accuracy of different counting methods. On at least ten events, different methods will be used and compared with the actual number of visitors present. This research has already taken place at the test conference of VISITFlanders in Antwerp and at the weekly market in the municipality of Bornem. This applied scientific research, with the support of the Flanders Innovation & Entrepreneurship, will result in an online tool that advises the appropriate counting method for each event, with a clear insight into the error margin per counting method.

Whether the counting of visitors at events is even more important now than before the corona crisis? "Absolutely", says Christine Merckx - head of the centre of expertise Public Impact. "Before the corona crisis, counting was mainly important for commercial and economic reasons." And then came the corona crisis ... "In times of corona, it is even more important for safety reasons. Counting must be done everywhere to guarantee the corona and distance measures. That is impossible if there is no accurate picture of the number of visitors."


Christine Merckx at the test conference in Antwerp on the 27th of April, 2021.


Towards one sole golden standard for visitor counting?


For the construction of the technological counting installations on the market of Bornem, the research team already arrived at 6 o'clock in the morning, together with the first market vendors. Cameras and trays were swiftly hung up in strategically chosen places. Metres of cabling and wires were rolled out. A crew room filled with hardware was set up to monitor the data during the market. Many screens filled with data dance to the rhythm of the visitors during the test case. When asked whether there is a counting method that does indicate exactly how many people are at a certain place, University of Antwerp researcher Raf Berkvens answers somewhat disappointedly: "Unfortunately not. From previous research, in which we developed new counting methods, we discovered that there is not really one sole golden standard."


By using test events, the research team wants to compare the different counting methods. Raf Berkvens explains: "The goal is to test the counting methods at least at ten different types of events of different sizes. Each time, we will use the different technologies. Then the question is: is the smallest number of visitors the right one? Or is the other technology more accurate at any given time? For each technology, and depending on the specific variables of the event, we calculate the margin of error. And we use that as a calibration."


Raf Berkvens at the market in Bornem on the 12th of April, 2021.


The importance of test events


This research cannot be conducted without test events. However, due to the corona crisis and the lockdown of the event sector, selecting suitable events for this research suddenly became a huge challenge. Then came the liberating news that the government would organise or commission test events in order to reopen the sector as safely as possible. The first official test conference in Flanders was organised by VISITFlanders. On the 27th of April, the FMCCA in Antwerp filled up with 200 event professionals. After each visitor was subjected to a corona quick test, they were counted in five different ways. Following methods were used: click count, quadrant count, Wi-Fi count, radio wave count and camera count.


"This kind of test event makes it easy for us to compare technologies because everything is very structured. There is a huge organization and a huge flow that is fixed. This allows us to compare our data much better. It is also an honour to be able to participate in this first official test event in Belgium. And especially to be able to back it up with scientifically based use of technology," explains Maarten Weyn, professor at the University of Antwerp. "For our research project, we normally count with mobile data as well. But this test conference does not lend itself to this. Counting with mobile data is more suitable for larger events with a larger number of visitors."


Maarten Weyn at the test conference in Antwerp on the 27th of April, 2021.


The development of an online Counter for events


The outcome of our research will result in a free online tool. The envisaged Counter-tool will help the various players in the event sector to make a well-founded choice. By entering a number of parameters - such as the surface of the event site, the number of access roads and the available budget - the event professional gets a clear view of the suitable counting methods for that event, with a clear view of the error margins per counting method, how much it costs and where it can be acquired.


In the summer of 2022, the final version of the online tool, with insights into the accuracy of each of these counting methods, is expected. Can't wait? Want to gain more insight into the suitable counting methods for your event today? Then you can already use the first version of the Counter.


With the support of the Flanders Innovation & Entrepreneurship.