To be able to estimate physics parameters for a battery under usage conditions a ML approach can be used. This in return requires a lot of data to simulate aging paths and also to generate raining data from these aging paths. Simulations of these needs to be done in such a way that one can generate: 1. Enough aging trajectories to calculate a good representative covariance matrix 2. Training dataset for a NN based upon aged trajectories and covariance To generate the aging trajectories a cell needs to be simulated for a whole lifetime and data of interest needs to be stored at various resolutions during charging/discharging cycles. These trajectories are intended to be created using a simulation package called PyBaMM to store time series data at a 0.1 second cadence for a specific drive cycle discharge and a specific identification current profile. Future work also includes optimization of the dynamic identification current.