AI-Enhanced Edge Computing Framework for Battery Thermal Management in Last-Mile Electric Vehicle Fleets
DOI:
https://doi.org/10.56472/WCAI25-130Keywords:
Battery thermal management, last‑mile delivery, Electric vehicles, Edge computing, Model predictive control, long short‑term memory (LSTM), Digital twinAbstract
The electrification of last-mile delivery fleets raises operational risks associated with the thermal behavior of lithium‑ion batteries due to environmental conditions that may include elevated ambient temperature, stop‑and‑go duty cycles, frequent HVAC transients when the door is opened, and intermittent fast charging. This manuscript describes an AI‑enhanced, edge‑computing framework for the real‑time prediction and control of battery temperature in commercial electric vehicles. The approach combines a control‑oriented resistance–capacitance (RC) thermal model with a long short‑term memory (LSTM) forecaster and a model predictive control (MPC) policy to maintain temperatures within safety limits while minimizing auxiliary energy use and computation latency. A layered vehicle edge compute and cloud architecture is specified to support sub‑100 ms control loops, privacy‑preserving data handling, and fleet‑level learning. The methodology includes (i) derivation and discretization of the RC model, (ii) physics‑informed training of the LSTM forecaster, and (iii) MPC formulation with actuation and charging constraints; practical design choices for embedded deployment (quantization, scheduling, watchdog fallbacks) are also detailed. A hypothetical case study for 100 last‑mile vans operating at 40 °C ambient conditions is outlined with evaluation metrics for thermal safety, HVAC energy per kilometer, operational uptime, and end‑to‑end loop latency on Jetson‑class systems. The proposed framework enables rigorous, data‑ready pilots even without proprietary telemetry by leveraging digital‑twin synthesis and synthetic route perturbations. The work aims to inform scalable, safety‑conscious electrification strategies for last‑mile fleets
Downloads
References
[1] J. C. Ferreira and M. Esperança, “Enhancing sustainable last-mile delivery: The impact of electric vehicles and AI optimization on urban logistics,” World Electr. Veh. J., vol. 16, no. 5, Art. no. 242, Apr. 2025. https://doi.org/10.3390/wevj16050242
[2] M. Hyland and D. Yang, “Electric vehicles in urban delivery fleets: How far can they go?,” Transp. Res. Part D: Transp. Environ., vol. 129, Art. no. 104127, Apr. 2024. https://doi.org/10.1016/j.trd.2024.104127
[3] Y. Ma, H. Ding, H. Mou, and J. Gao, “Battery thermal management strategy for electric vehicles based on nonlinear model predictive control,” Measurement, vol. 183, Art. no. 110115, Sep. 2021. https://doi.org/10.1016/j.measurement.2021.110115
[4] Q. Hu, M. R. Amini, A. Wiese, J. B. Seeds, I. Kolmanovsky, and J. Sun, “Electric vehicle enhanced fast charging enabled by battery thermal management and model predictive control,” IFAC PapersOnLine, vol. 56, no. 2, pp. 10684–10689, 2023. https://doi.org/10.1016/j.ifacol.2023.10.721
[5] R. Wang, H. Zhang, J. Chen, R. Ding, and D. Luo, “Modeling and model predictive control of a battery thermal management system based on thermoelectric cooling for electric vehicles,” Energy Technol., vol. 12, no5. 2024. https://doi.org/10.1002/ente.202301205
[6] L. Liu, G. Xu, Y. Wang, L. Wang, and J. Liu, “Battery temperature estimation at wide C-rates using the LSTM model based on polarization characteristics,” J. Energy Storage, vol. 84, Art. no. 113941, Sep. 2024. https://doi.org/10.1016/j.est.2024.113941
[7] J. Han, J. Seo, J. Kim, Y. Koo, M. Ryu, and B. J. Lee, “Predicting temperature of a Li-ion battery under dynamic current using long short-term memory,” Case Stud. Therm. Eng., vol. 63, Art. no. 105246, Nov. 2024. https://doi.org/10.1016/j.csite.2024.105246
[8] Wang, XT., Wang, JS., Zhang, SB. et al. Capacity prediction model for lithium-ion batteries based on bi-directional LSTM neural network optimized by adaptive convergence factor gold rush optimizer. Evol. Intel. 18, 35 (2025). https://doi.org/10.1007/s12065-024-01013-7
[9] Z. Zhu, Y. Zhang, A. Chen, J. Chen, Y. Wu, X. Wang, and T. Fei, “Review of integrated thermal management system research for battery electrical vehicles,” J. Energy Storage, vol. 84, Art. no. 114662, Dec. 2024. https://doi.org/10.1016/j.est.2024.114662
[10] A. Alawi, A. Saeed, M. H. Sharqawy, and M. Al Janaideh, “A comprehensive review of thermal management challenges and safety considerations in lithium-ion batteries for electric vehicles,” Batteries, vol. 11, no. 7, Art. no. 275, Jul. 2025. https://doi.org/10.3390/batteries11070275
[11] J. S. Menye, M.-B. Camara, and B. Dakyo, “Lithium battery degradation and failure mechanisms: A state-of-the-art review,” Energies, vol. 18, no. 2, Art. no. 342, Jan. 2025 https://doi.org/10.3390/en18020342
[12] H. Wei, C. Callegari, A. C. O. Fiorini, R. Schaeffer, and A. Szklo, “Technical and economic modelling of last-mile transport: A case for Brazil,” Case Stud. Transp. Policy, vol. 14, Art. no. 101219, Jun. 2024. https://doi.org/10.1016/j.cstp.2024.101219
[13] Aragani, Venu Madhav and Maroju, Praveen Kumar and Mudunuri, Lakshmi Narasimha Raju, “Efficient Distributed Training through Gradient Compression with Sparsification and Quantization Techniques” (September 29, 2021). Available at SSRN: https://ssrn.com/abstract=5022841 or http://dx.doi.org/10.2139/ssrn.5022841
[14] Sandeep Rangineni Latha Thamma reddi Sudheer Kumar Kothuru , Venkata Surendra Kumar, Anil Kumar Vadlamudi. Analysis on Data Engineering: Solving Data preparation tasks with ChatGPT to finish Data Preparation. Journal of Emerging Technologies and Innovative Research. 2023/12. (10)12, PP 11, https://www.jetir.org/view?paper=JETIR2312580
[15] Sehrawat, S. K., Dutta, P. K., Bhatia, A. B., & Whig, P. (2024). Predicting Demand in Supply Chain Networks With Quantum Machine Learning Approach. In A. Hassan, P. Bhattacharya, P. Dutta, J. Verma, & N. Kundu (Eds.), Quantum Computing and Supply Chain Management: A New Era of Optimization (pp. 33-47). IGI Global Scientific Publishing. https://doi.org/10.4018/979-8-3693-4107-0.ch002
[16] S. Panyaram, "Digital Transformation of EV Battery Cell Manufacturing Leveraging AI for Supply Chain and Logistics Optimization," International Journal of Innovations in Scientific Engineering, vol. 18, no. 1, pp. 78-87, 2023.
[17] Mohanarajesh, Kommineni (2024). Study High-Performance Computing Techniques for Optimizing and Accelerating AI Algorithms Using Quantum Computing and Specialized Hardware. International Journal of Innovations in Applied Sciences and Engineering 9 (`1):48-59.
