The growing integration of electric vehicles and renewable energy systems has heightened the need for efficient and reliable battery management systems (BMS). Accurate state-of-charge (SoC) estimation is a crucial element in maintaining battery performance and longevity, yet conventional model-based and neural network approaches often struggle to capture complex and nonlinear battery dynamics. To address these limitations, this study presents a proof of concept based on a Reservoir Spiking Neural Network (RSNN) designed to improve SoC estimation accuracy through biologically inspired computation. The proposed model combines the temporal processing strength of spiking neural networks with the dynamic memory of reservoir computing, enabling the network to effectively represent temporal dependencies in battery discharge patterns. Different spike encoding strategies were explored to optimise the transformation of continuous battery data into spiking activity. The analysis reveals that population-based encoding enhances prediction accuracy by enriching the internal spiking representation within the reservoir, though it requires slightly higher synaptic activity. Experimental results demonstrate that the RSNN provides competitive SoC estimation performance during critical discharge stages, without a significant increase in training time. These findings establish RSNNs as a promising alternative for SoC estimation, offering improved adaptability, high computational efficiency, and suitability for low-power embedded implementations in advanced BMS applications.