​Drivers always try to plan their trips over the road network by avoiding highly congested areas. Several research studies have considered image processing techniques and vehicular network technologies to predict the level of traffic congestion over pre-determined road scenarios. In this paper, we aim to use big data and machine learning regression techniques to provide the driver ahead of time with the expected level of traffic congestion on the investigated road scenarios. Three basic regression techniques have been experimentally tested over some main roads in Amman, Jordan. Fast and correct traffic characteristics have been predicted on the investigated road scenarios, such as positions, speed, and locations of each vehicle. After that, the best regression technique has been used to predict the level of traffic congestion on the traversed road network during future periods. These predictions are also reported to the moving traffic to be considered in the route options of each moving vehicle.​