Porous media systems are relevant in many research fields of chemical engineering: packed bed catalytic reactors, filters, and batteries. Microscale modeling (i.e., pore-scale) of a representative volume of the porous structure is a state-of-the-art methodology to accurately evaluate transport-related properties, such as reaction rates or filtration efficiencies. Computational fluid dynamics (CFD) has been widely employed to this end. However, these microscale simulations are computationally expensive, making it difficult to integrate them into multiscale modeling or optimization workflows where fast-response models are needed. Machine learning, particularly deep learning techniques, can be used as effective and accurate digital twins. Convolutional neural networks (CNNs) are appropriate models for porous media applications because they can take images of the porous structure as inputs, allowing the network to select the most relevant geometrical features automatically. In this work, we train multiscale convolutional neural networks (MSNet) [1] as surrogate models for the local prediction of fields in porous media, focusing on filters and the discharge of lithium-ion batteries. The workflow we propose involves creating a dataset from CFD-based simulations (performed with OpenFOAM), which is then used to train the neural networks and test their generalization capabilities. For the filtration application, CFD simulations have been conducted on a wide range of geometries and operating conditions. These serve as the input features for MSNet, which is trained to predict the concentration fields, achieving errors on the effluent concentration of less than 5% [2]. For the lithium-ion battery case study, discharge simulations of the half-cell (cathode side) were solved with COMSOL. To handle the transient nature of the dataset, MSNet was modified with an autoregressive approach. The resulting model accurately predicts lithium concentration and potential dynamics in the solid phase. [1] Santos, Javier E., et al. Transport in porous media 140.1 (2021): 241-272. [2] Marcato, Agnese, et al. Chemical Engineering Journal (2022): 140367