The recent drive to decarbonise transport has prompted a sharp rise in the global sales of electric vehicles (EVs). Charging infrastructure provides an interface between the transport and energy sectors and enables the integration of EVs into the electrical network. The transition to EVs requires huge investments into new public and private charging infrastructure to meet the mobility demands of the population. To maximise their utilisation, new charging infrastructure will need to be installed in areas of high existing and future demand. Furthermore, as the adoption of EVs continues to grow, at peak times the energy demand due to charging could surpass the capacity of local transformers, thus requiring expensive grid upgrades. As we move towards a more distributed energy system, emerging players, such as EVs, are beginning to form vast networks with complex interactions. Modelling the interactions between EVs and the grid is made challenging by the following factors. Firstly, there is a large amount of technological diversity within the range of available EVs and chargers, making it difficult to accurately represent the whole system. Secondly, the introduction of smart charging functionalities, where the charging time and power is shifted based on grid requirements and the vehicle owner's needs, introduces more potential flexibility and degree of control over charging. Thirdly, the battery charging dynamics dictating the charging demand varies between EV models and is sensitive to the battery state-of-charge. Finally, the movement and charging patterns of EVs are dictated by human behaviour, which is often difficult to predict and rationalise. This complexity means that models used to capture the aggregated dynamics of the system tend to rely upon simplified representations that often deviate substantially from reality. Over-simplification of these dynamic interactions can potentially have significant consequences. For example, it could lead to a mismatch of energy supply and demand, uneconomic grid development, and ultimately delayed and/or high cost in system decarbonisation. Current attempts to model and optimise EV charging tend to make unrealistic assumptions about consumer travel and charging behaviour. Moreover, there has been limited spatiotemporal modelling of EV-grid interactions that accurately captures the grid topology at different scales. In addition, attempts at long-term demand forecasting often do not accommodate for changes in EV and charging technology, as well as user behaviour. This PhD aims to answer the following research questions: 1) What will future EV charging demand look like, and how will it respond to evolving technology and user behaviour? 2) What is the current and future impact of EV charging on the grid, and where might reinforcement be required? 3) Where should new chargers be installed to meet people's mobility needs and reduce the need for grid reinforcement? The goal of this research is to develop a high-fidelity spatiotemporal model to build a better understanding of the current and future aggregated group dynamics of EV charging demand. This will be achieved using a combination of data-driven and agent-based modelling techniques. The key parameters that influence this aggregated demand should be identified to inform the spatiotemporal modelling, optimisation, and interactions between the charging and grid infrastructure. The research should inform the future infrastructure rollout strategy of network operators and local councils. The aim is that this research will enable cheaper and faster integration of EVs into the grid and thus accelerate the decarbonisation of the automotive and power sectors.