The current restrictions on the registration of combustion engines in different countries and the harmful impacts of fossil fuels on the environment and human health have motivated decision-makers to use batteries and/or fuel cells as alternatives for combustion engines in different applications. Although there has been considerable progress to commercialize batteries in the automotive industry, which demands an average range of 300 km, the low range and high weight of batteries has prevented them to be an option for maritime and aviation applications. On the other hand, proton exchange membrane fuel cells (PEMFC), known as the most commercialized type of fuel cell for mobile applications, benefit from high ranges. The main aim of this thesis was to investigate PEMFC on different scales (micro-, meso-, and macro- scale). In particular, the role of computer science using artificial neural networks has been considered. Thermal management has been treated as a shared topic in the various analyzed scales. In the micro-level studies, the presence of capillary pressure results in the creation of water channels in the GDL/MPL, called capillary fingering. This phenomenon fills the pores of those layers and reduces cell performance by flooding. That is why an analysis of the water distribution is required to understand and limit the operating window of PEMFCs. In this regard, a simulation model has been developed to analyze the impact of changes in the GDL contact angle, porosity, and permeability on the GDL liquid water removal, which has a direct relation with the water/thermal management of PEMFCs. Focused Ion Beam- Scanning Electron Microscopy and Computational Tomography (CT) scans were used in this scale for deepened analysis. The main objective of the meso-level study was to explore the potential for performance improvement of PEMFCs by the implementation of porous media in the gas flow channel. The effect of this layer on thermal/water management was evaluated using simulations with ANSYS software considering measurable PEMFC metrics such as voltage, power density, Nusselt (Nu) number, and pressure drop. The synergistic de-sign of the flow channel and the structure of the porous layer in the channel are key to improving thermal/water management, which was evaluated through simulations. A new parameter called Evaluation Criterion of Proton Exchange Membrane (ECPEM) was introduced using ANN modeling to optimize the performance of the system considering the voltage, power density, Nusselt (Nu) number, and pressure drop. In the macro-level studies, the goal of this thesis was to analyze the possibility of different electrochemical technologies to provide the required power for different purposes. The system-level analyses are mainly developed with PEMFC for mobile applications, while Solid Oxide Fuel Cells (SOFC) have been suggested for stationary purposes. In mobile applications, the use of PEMFC has been analyzed for shipping and flight. Lithium-Ion (Li-Ion) batteries were combined with PEMFC and dynamic performances have been developed. It was shown that the integration of Li-Ion batteries and PEMFC can be considered as a solution for shipping applications with around 1 MW of power to transport 780 passengers for short distances (ferry). The use of integrated fuel cell-battery systems in flight applications was illustrated with a 14 kg drone in three different power demand profiles.