Climate change and the diminishing availability of fossil fuels necessitate a shift toward sustainable and renewable resources. This transition is evident in the development of low-CO2 emission hybrid and electric vehicles. As our dependence on these vehicles grows, energy storage systems, particularly electrical energy storage systems like batteries and supercapacitors, play a crucial role. While Li-ion batteries excel in performance, with energy densities exceeding 200 Wh/kg, supercapacitors are gaining attention for their rapid power delivery in applications where Li-ion batteries fall short. However, the low energy density of supercapacitors poses a challenge, prompting the need for innovative materials or charge storage concepts. The use of nanoporous carbon electrodes has significantly increased capacitance, and molecular modeling, coupled with in situ experiments, is essential for understanding the underlying molecular phenomena. Implementing multi-scale models, especially in the context of the MultiXscale project, can enhance predictive capabilities, allowing for simulations at larger scales and closer approximation to experimental conditions. The project aims to develop efficient workflows for sequential and concurrent coupling of different scales, utilizing machine learning potentials to bridge the gap between quantum mechanical accuracy and numerical efficiency. This approach will be applied to model a microsupercapacitor, exploring the impact of geometric design. Additionally, the same methodology could be extended to study redox reactions and proton transport in fuel cells, addressing the challenges of chemical reactions, multiphase flows, and complex topologies through proper mesoscale modeling and code coupling.