Green Wave today

About a year ago I first read about Travolution , an experiment in transmitting information from a traffic light to a car. In short, the idea of ​​the experiment is to prompt the driver when the traffic light switches and the recommended speed in order to get into the "green wave", that is, move without stopping at traffic lights. About this experiment there was an article on Habré . The results of the experiment were not said to be astounding, but positive. But despite this, there is no talk of practical application of this technology.
Obviously, the reason for this is the complexity of the implementation, because in order for the system to work, we also need specially equipped traffic lights and specially equipped cars, that is, a radical change in infrastructure.

But recently I came up with the idea of ​​how to bring the "green wave" to life with the existing infrastructure.

In my view, three key components are needed to implement this technology:

1. A tool that would visualize the data on the switching time of the traffic light to the driver.
2. A way to get data from a traffic light.
3. A way to transfer this data to the first tool.

Obviously, the data visualizer is a smartphone. I understand that not everyone has a smartphone yet, but this is all going on.

The data transfer method is mobile internet. And yes, here, too, not everything is smooth, and the coverage is not everywhere, but again, there is progress in this direction.
')
But how to get data from a traffic light?

Option 1 is fantastically expensive for the CIS and quite real for foreign countries:

Agree with the authorities and put the sensors in traffic lights so that they transmit data on the work of the traffic light to the data center via the mobile Internet.

Option 2 - more or less real in the CIS:

Agree with companies owning buildings that are located near the desired traffic lights, and, if they have external video surveillance, agree that video from a single camera, which looks just at the right traffic light, is transmitted to the data center.
If there are no buildings with video cameras nearby, then you can arrange to install a standalone IP camera, for example, such that it transfers its data to the data center. Yes, it can not be called "using the existing infrastructure," but it seems to me that this is quite realistic, even on a large scale.
In the datacenter you will need to recognize the traffic light on the received image and what color it is lit. For a picture from a still camera, the task is completely trivial. And even to transfer to the data center the whole picture is not needed, you can only a small area with a traffic light.

No matter how the data about the switching time of the traffic light were obtained, they are transmitted to the smartphone. But to transfer data via the Internet in real time will not work. There will inevitably be some kind of delay. Therefore, the smartphone should be transmitted data on the time of previous switching traffic lights. On the smartphone, the application, based on the incoming data and the data set of previous traffic light switches, should predict the time of the next traffic light switch and bring this information to the driver.

Of course, all these predictions will only work for ordinary traffic lights, which switch at regular intervals and the switching mode of which changes only a few times a day, if at all. That is, at smart traffic lights, such a system is useless. Especially if the data will be shot through the camera. But, "fortunately" in the CIS countries there are almost no smart traffic lights. If in Moscow there are about 3 thousand traffic lights, of which only 50 are managed , then in the smaller cities there are probably even fewer smart traffic lights.

Well, in countries where there are smart traffic lights, firstly, it’s quite realistic to agree on taking data directly from traffic lights, and secondly, you can take data not even from traffic lights, but directly from the traffic control center. Moreover, this way it is possible not to predict the switching time of traffic lights, but to take real data on the next planned switching of traffic lights from ASUDD data centers .

After that, these data need to be visualized and displayed on the smartphone screen.
Here I came across this idea:

“It is proposed to install a strip of lamps or gas discharge tubes along the highway, which will glow in green or red light on the sections. Luminous areas should move along the highway at a speed equal to the speed of the "green wave". It’s enough for the driver to maintain such speed that he is in the green zone. ”

It seems to me that this is a very good idea for visualizing the “green wave”.
Applying this idea to modern realities, we get a picture from the post title.

Road marking by zones:

• the red part is the area where cars face the traffic lights (red and yellow traffic lights);
• yellow-green is the front, risky part of the "green wave", where there is a risk of stopping in front of cars standing at the traffic lights.
• green - the back of the "green wave", where the most profitable ride.

Other designations

• The red hexagon is the time to switch the nearest traffic light.
• In the green circle, the recommended speed to continue driving in the green wave.
• The blue triangle is a vehicle.

I believe that such a service is not only real with the existing infrastructure, but may also be beneficial for the developer company that would take it. But a small company can not do it. Too big investments are needed in the creation of infrastructure for collecting information on the response time of traffic lights. But some big company specializing in navigation services, or simply having, among others, its own navigation service, may well put this idea into practice and stand out noticeably from other software vendors for in-car navigation.

And it seems to me that this is a really useful service, for which it is not a shame to ask for money. Of course, this system will not work during peak hours, but at other times, it can be really useful and convenient. It would be especially useful for buses in cities where there is a dedicated line for public transport.

PS For Moscow, the system is completely useless, in any form.

Update1: Based on a comment from the user izmalk .
The system may be potentially dangerous, since indeed, the forecast of the application about the switching time of the traffic light may not come true. Therefore, the front of the green wave should be pushed back to the braking distance for this car at a given weather and at a given speed, multiplied by the safety factor. And then the application will show information where it is safe to go in the “green wave”. The question is whether something will remain of the green wave at the same time.
And yes, for this you will need to also transfer weather information to the device.

Based on a comment from the user inborn_killer .
You can also enter some component of social responsibility in the application.
If the driver uses the application to jump through the red wave and get into the green with overspeeding, the application is disabled for a while, punishing the user for the excess. On the one hand, this is “negative feedback” and this is usually a taboo for developers, but on the other hand, it can save lives. Yes, and when speeding, the stopping distance increases, so we can assume that the "green wave" is simply eaten by the braking path.