Time dependent origin to destination, OD, matrices are the key input to dynamic traffic models, mainly to simulation models, microscopic as well as mesoscopic. Dynamic Traffic Models, DTM, are one of the major components of the Advanced Traffic Management Systems and Advanced Traffic Information Systems. DTM play a crucial role in estimating the current traffic state and forecasting its short term evolution and the quality of the results that they provide depends, not only on the quality of the models, but also on the accuracy and reliability of the inputs and, therefore, on the quality of the dynamic or time dependent OD matrices that are part of that input. The research reported in this paper explores two complementary issues for estimating OD matrices: the exploitation of travel time measurements provided by sensors detecting Bluetooth devices equipping vehicles (Tom-Tom, Parrot, hands free...) which define the inputs to an enhanced Kalman Filtering model; and the use of data supplied by V2I technologies (i.e. positions and speeds) that allow the estimates of direct samples which combined with a path reconstruction process provide the desired OD estimates.