Optimization of how citizens are distributed on the wagons of a Brussels Metro

Dear Brussels Smart City team.

I couple of years ago, I went to Sao Paulo in Brazil (a city about 15 times bigger than Brussels) and while taking the metro I saw something great, something that I had never seen before. Since that that I have always asked myself: Why don't we have that thing in Brussels?

They have in place a very simple system that helps A LOT to better distribute the people on a metro, with this they avoid having some wagons extremely crowded while some others have much more space. It works like that, if I remember correctly :)

• The metro will come with X number of wagons, let's say, 10. Some of these 10 wagons are much more crowded while others are less busy. They allow people to have an idea of how busy the wagons are BEFORE the metro arrives, people can know what are the less busy wagons 1~2 mins before.
• How exactly they do it? They put some kind of traffic lights on the metro stations, one small traffic light for each wagon. For our example it would be 10 small traffic lights on a metro station. To keep it simple, each traffic light will have GREEN (Lot of space on the wagon, please take this one), ORANGE (there is still some place, it should be Ok), or RED (the wagon is completely full, please go to another wagon). Simple, right? So in the practice, you see that there much less people on the red lights (which helps not only to the passengers who wanna take the metro but ALSO helps the passengers who wanna leave the metro because the have much space to walk out of it). As I said, for the users it's extremely simple to understand, simple to use and simple to see the Huge benefits.
• Now, how does it actually work? Well, we will need to have somehow a quite accurate idea of how busy the wagons are. I don't know exactly work the system in SaoPaulo but here we could make it work counting the people like that: Counting the "heads" or counting the "faces". I know there is already existing software that recognizes when there is a face so we don't need to re-invent the wheel. Also, we don't actually need to recognize A face (that's not what we want), what we need to recognize is a quite accurate number of faces, and I think that should be feasible. Then, for each wagon, we would need to select a color (Green, Orange or Red) depending the aprox. number of faces found on a wagon and .. .. .. that's it !

Also: Keeping in mind the idea of social distancing and so on, this project could be very helpful !