What facts can be extracted from the video stream of outdoor surveillance cameras?

Question

Video

The provider has many public surveillance cameras in the city. I have not yet come up with anything other than “determining whether the fountains are open after winter”, and “how many cars are in the queue at the gas station”. What other facts can be obtained using computer vision?

Answer 1

How many people pass under the camera, is the lantern in the field of view working, the presence of snow, cloudiness, seasonal changes in leaf cover, some weather phenomena (snow, fog).

Answer 2

Count passing traffic in public places and sell it to landlords in the vicinity. The issue is that the cameras are not installed based on this and it is problematic to automate.
Here Uncle Po talks about this topic.
https://t.me/kurspo/600