Hybrid and Low-Power AI Techniques for Battery Management Systems in Electric Scooters: A Literature Review

This paper presented a focused review of battery management system approaches for electric scooter applications, emphasizing the evolution from traditional estimation techniques to modern hybrid and low-power artificial intelligence methods. Traditional, machine learning, and deep learning approaches were discussed to provide the necessary background and highlight the increasing adoption of data-driven techniques for battery state estimation.   Particular attention was given to hybrid approaches, including CNN--Transformer, CNN--LSTM, Transformer--LSTM, EKF--LSTM, LSTM--GRU, LSTM--Random Forest, Physics-Informed Neural Networks, and Attention--LSTM models. The review showed that hybrid architectures generally achieve superior prediction performance by combining complementary learning mechanisms, filtering techniques, and domain knowledge. However, these improvements are often accompanied by increased computational and memory requirements.   Low-power approaches such as TinyML and quantized neural networks were also examined as promising solutions for deploying intelligent battery management algorithms on resource-constrained embedded platforms. The comparative analysis highlighted a clear trade-off between estimation accuracy and deployment efficiency. While hybrid models offer the highest prediction accuracy, low-power techniques provide practical deployment advantages for real-world battery management systems.   Future battery management systems are expected to benefit from the integration of advanced hybrid architectures with lightweight optimization techniques, enabling accurate, efficient, and scalable solutions for next-generation electric scooters and other electric mobility applications.

[1] A. A. Dhamangaonkar and R. P. Hasabe, “Evolving battery management systems for ev,” in 2025 4th International Conference on Computational Modelling, Simulation and Optimization (ICCMSO), 2025, pp. 298–303. [2] X. Zhang, X. Zhang, and W. Zhang, “Current status of research on factors causing lithium-ion battery packs performance degradation and their state of health estimation,” in 2023 5th International Conference on Power and Energy Technology (ICPET), 2023, pp. 306–311. [3] S. A. Hasib, S. Islam, R. K. Chakrabortty, M. J. Ryan, D. K. Saha, M. H. Ahamed, S. I. Moyeen, S. K. Das, M. F. Ali, M. R. Islam, Z. Tasneem, and F. R. Badal, “A comprehensive review of available battery datasets, rul prediction approaches, and advanced battery management,” IEEE Access, vol. 9, pp. 86 166–86 193, 2021. [4] G. Fathoni, S. A. Widayat, P. A. Topan, A. Jalil, A. I. Cahyadi, and O. Wahyunggoro, “Comparison of state-of-charge (soc) estimation performance based on three popular methods: Coulomb counting, open circuit voltage, and kalman filter,” in 2017 2nd International Conference on Automation, Cognitive Science, Optics, Micro Electro-Mechanical System, and Information Technology (ICACOMIT), 2017, pp. 70–74. [5] A. Lotfivand, D. Yu, and B. Gomm, “State of charge estimation using extended kalman filter in electric vehicles,” in 2022 27th International Conference on Automation and Computing (ICAC), 2022, pp. 1–5. [6] K. Xiao, Z. Cui, and N. Cui, “State-of-charge estimation for lithiumion battery pack based on unscented kalman filter considering cell inconsistency,” in 2022 6th CAA International Conference on Vehicular Control and Intelligence (CVCI), 2022, pp. 1–6. [7] S. Vattem and S. R. Gorantla, “A critical review on available methods for estimating the present state-of-charge of the batteries used in ev/hev,” in 2023 International Conference on Advanced & Global Engineering Challenges (AGEC), 2023, pp. 26–31. [8] S. K. Sunori, S. Jain, P. Juneja, and A. Mittal, “Implementation of machine learning in battery management system of electric vehicles,” in 2024 2nd World Conference on Communication & Computing (WCONF), 2024, pp. 1–5. [9] B. R. Kumar, S. N, M. U, S. M, and S. K, “An ann - deep learning based prognostic framework for electric vehicle battery health management system,” in 2026 5th International Conference on Communication, Computing and Electronics Systems (ICCCES), 2026, pp. 417–422. [10] U. Gezer, M. Kuc¸¨ uk, O. Yazar, S. Coskun, and F. Zhang, “A hybrid cnn- ¨ lstm framework for state of charge estimation of lithium-ion batteries,” in 2025 9th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT), 2025, pp. 1–6.