you can try this API:

https://developer.here.com/documentation/traffic/topics/additional-parameters.html

this will give you flow in a JSON string for the interested area.

you can also use google API Direction which is not free but is more reliable.

have fun!

@Mouna, thank you but this is not actually what I am looking for. I know how to build a scenario. But I need to get real data from somewhere and pass it to SUMO to generate the route file.

@Ali, thanks for your answer.. I will check that.

Well I can answer the second part of your question: It depends on the data type you are able to gather for this study. One way (and only that worked for me) to do it is to use Dynamic Calibrator module (http://sumo.dlr.de/wiki/Simulation/Calibrator#Calibrators). Data required for this process are traffic counts located somewhere on the network.

First, you need to generate random trips using randomTrips.py script (http://sumo.dlr.de/wiki/Tools/Trip#randomTrips.py) that creates XML files with routes and trips.

After that, you need to create calibrators, accordingly to traffic counter positions on the network (using a calibrator.xml file from Calibrators page above). I've used python code since I had hundreds of thousands of readings.

Finally, you run the simulation. Calibrators adjust the vehicles flow on counter positions, so the overall number is approximately close to the real readings.

@Kamil

Thank you. This is exactly what I need.

Hi Rehab Shaheen and Kamil Smolak

were you successful with this?

Kamil Smolak should the file calibrators.xml be manually generated?

Hi Toufik Khawli,

The file "calibrators.xml" was generated through self-made python code. It writes the structure of the XML file from the database I had. Data used for generation were counts and link IDs (everything was in the database).

Calibrators were adding or removing cars from links according to the readings from counters. The more calibrators you have the more accurate the simulation is.

Thanks Kamil Smolak for the reply.

I was able to manage it somehow, after choosing a location on a map for the calibrator then tuning the vehicleperhour parameter (works fine so far).

However I have noticed that this parameter affects the total number of cars in the simulation which is controlled by the parameter "-p" (when generating the random trips). I guess some kind of scaling between these parameters need to be taken into consideration.

Toufik Khawli Great to hear it is working now.

In my opinion "-p" should be scaled in a way that it will satisfy the total number of cars that should be existent in the network in a reasonable time. One idea is to set it to the value that correspond to the "network filling speed", for example when people are leaving their houses in the morning.

Good Answer Kamil Smolak

Kamil Smolak Thank you for your answer. Can you please provide me with your scripts. The link in your comment above does not work for me. Thank you