To extend the usage condition and prevent batteries from accelerated aging battery models used in battery management systems will need to be more advanced. One approach is to use electrochemical models such as Duller-Fuller-Newmann (DFN) model for estimations to access internal battery states. These internal states could perhaps be used to enable better charging/recharging capabilities without damaging the battery in long run and hence optimize the usage of existing batteries. This model is difficult to parametrize to begin with but if done well good accuracy can be achieved. However, when a battery is cycled / used it will naturally start to degrade and the initial model will lose accuracy of important internal states. This project aims to increase the accuracy overtime of the DFN model by utilizing ML to re-parameterize aging related parameters.