Remaining Useful Life (RUL) estimation of Lithium-Ion batteries is rapidly gaining significant attention worldwide. In a recent work a hybrid approach has been introduced aiming at improving the accuracy and precision of batteries prognostic merging aspects typical of condition monitoring techniques, Kalman and Particle filtering methods (i.e. the state space of the battery) and artificial intelligence estimation by means of Recurrent Neural Network (RNN). The proposed method uses the state space of the battery to effectively train the RNN according to a new single exponential model. Building upon this innovative approach, this paper aims at validating the proposed method testing its performances on an additional battery dataset provided by Toyota Research Institute in partnership with Massachusetts Institute of Technology and Stanford University. The considered dataset includes over one hundred batteries tested under different fast-charging conditions. The results achieved in this work for different batteries and different charging policies highlighted the remarkable ability of the proposed method to precisely estimate the RUL of batteries under several different conditions. This helps validating the proposed approach with the aim of extending its range of applicability without any restrictions.