You are here:

+90 (312) 210 00 15

The Nerd Genie and Traffic

Once upon a time, I came across an old rotten keyboard. I wiped it, and a nerdy genie came out!


Retrieved from AlternataivePress


Genie: Uyf, I was going to read a paper and here you bring me out… I hope it is worth it. You can ask me 3 questions and I will answer them then go back to my job.

Me: Wow a nerd Genie! let me start with my first question then. I am frustrated by stopping too long at junctions. I just wait for this stupid traffic light to come on! What do you suggest? please save my day!

Genie: Daym man, I never drove a car. But let me think… In old days, police personnel would stand on a podium wearing something shiny, with his/her hands waving to which road shall go/stop. …



Later on, they bring the idea of the traffic light with 3 lights; red for stop, yellow to get ready, and green to go. They call each combination of road openings a phase and some mechanism would switch between those phases based on predefined time.

Me: Lol, I asked you for a suggestion you gave me a history course of traffic intersections, so what solves my issue here?

Genie: hold on little boy,

Me: Don’t call me a little boy

Genie: Ok little boy, What I suggest and what really exists nowadays, is to have a top-level management system on top of the junction itself. Let me go into details: We can put some cameras there at the intersection. Wire them to a computer processing the incoming feed, it can count how many vehicles are waiting/passing and based upon that do the same logic that police there on the podium was doing.


Me: So what is the point?

Genie: Ah you dum dum! Now the time of stopping at red light won’t be static, it would change based on the situation out there.

Me: So, the magic box we call “a computer processing the incoming feed” can you tell me a bit more about it?

Genie: You owe me a meal for that! Now, with the recent development of technologies, we can run what they call “artificial intelligence” on this computer.

Here you go, something they call “supervised learning” in terms of computer vision, by basic terms is telling the computer that what it sees in a picture is -for example- a burger for thousands of times. Then not surprisingly the computer would recognize the burger when it sees it again…

The same logic can be applied here, you may pick up a “model”, give it thousands of vehicle pictures, and telling it that those are cars… I promise it won’t tell they are burgers!.


So, after the computer recognizes that car, it can count how many cars on that road, and decide the “phase” time to operate. Voila, you won’t wait a long time anymore!


Me: Wow, thanks genie. That is brief and nice. Genie: So what are the other 2 questions left? Me: Genie, let me ask you about traffic safety in the following episode!


**Note: The drawings may not represent the exact real world, they are just illustrations to basically understand the principle of interest


Mohamad is interested in technology, not more technical than the ethical, and future of humanity. Whether it is AI or a simple logical ladder. He enjoys writing about real-world trends in AI, sustainability, and mobility concepts. For more AI mobility concepts check out

Sosyal Medya'da Paylaş!