Driver scoring and risk analysis is gaining impact towards fuel economy and off-line driver's education. Today's cars are becoming computer controlled systems in which the driver's commands, car's motion and trajectories can be logged and transmitted to the control center by means of cellular communication, . Furthermore, local dynamic map involving the digital road map, speed limits and average link travel speed can be used to match the driving behavior with respect to the normal conditions of the static and dynamic traffic. On the other hand, driver scoring enforces better driving and operation of the vehicles in the fleets. Fleet managers can take corrective actions to the drivers according to their driving quality. In this paper, we define a safety oriented driving quality index and the corresponding architecture which is scalable up to 50.000 vehicles.