AI-OR Hybrid Models for Optimizing Electric Vehicle Fleet Management: Demand Forecasting, Routing, and Charging Station Allocation

Abstract

Managing a fleet of electric vehicles (EVs) comes with its own set of challenges—from anticipating fluctuating demand and monitoring battery health to ensuring efficient routing and charging. This paper explores a fresh approach that combines the power of Artificial Intelligence (AI) with traditional Operations Research (OR) optimization methods to tackle these issues in a practical way. I use AI techniques to predict key factors like how many vehicles will be needed and the state of their batteries, feeding these insights into sophisticated optimization models. These models then help plan vehicle routes and charging schedules in real time, aiming to reduce energy consumption and operational costs. Through a series of experiments using simulated data, my AI-OR hybrid model performed better than methods relying solely on optimization or AI, proving to be more effective at utilizing fleet resources and saving energy. What makes this approach especially promising is its potential to support more sustainable and cost-efficient transportation systems as electric vehicles become increasingly common. My findings provide a valuable roadmap for integrating predictive analytics with optimization in real-world EV fleet management, pushing the boundaries toward greener and smarter mobility solutions.

Introduction

The transition to electric vehicles (EVs) is transforming transportation and fleet management, offering environmental benefits such as lower emissions and reduced fossil fuel dependence. However, EV fleets face unique challenges, including battery degradation, limited charging infrastructure, and fluctuating demand, which complicate operational efficiency. Traditional fleet management methods are insufficient for these dynamic uncertainties, motivating the integration of Artificial Intelligence (AI) for demand and battery health forecasting with Operations Research (OR) optimization for routing and charging decisions.

This research develops a hybrid AI-OR framework that combines predictive modeling with stochastic optimization to manage EV fleet operations. AI forecasts enable accurate predictions of demand and battery state, while OR models dynamically optimize vehicle routing, charging schedules, and energy usage under uncertainty. Computational experiments using real-world and simulated data demonstrate that the hybrid approach outperforms purely AI-based or OR-only methods, reducing energy consumption, operational costs, and vehicle downtime while improving charging station utilization.

The study highlights the potential of integrated AI-OR methods for sustainable and efficient EV fleet management. By enabling adaptive, data-driven decision-making, the framework supports environmental goals, operational reliability, and cost-effectiveness. Limitations include reliance on data quality and computational complexity, suggesting future research directions in scalable algorithms, IoT integration, and renewable energy-aware fleet planning.

Conclusion

This paper presents a novel hybrid framework that seamlessly integrates Artificial Intelligence prediction models with Operations Research optimization techniques to enhance electric vehicle fleet management. By harnessing AI for demand forecasting and battery health estimation, and feeding these insights into stochastic routing and charging optimization models, the hybrid approach effectively tackles uncertainties and operational complexities faced by EV fleets. The computational experiments demonstrate that this integration leads to significant improvements in energy efficiency, cost savings, and vehicle utilization, outperforming traditional approaches that rely on AI or OR alone. These findings highlight the transformative potential of hybrid AI-OR models in enabling more sustainable, reliable, and cost-effective fleet operations. Looking forward, the adoption of such integrated frameworks can support the broader shift toward green mobility, helping fleet operators reduce their carbon footprints and operational risks. Practical applications range from urban delivery services to public transportation systems, wherever EV fleets are deployed. Future work will focus on enhancing scalability, incorporating real-time data streams, and expanding to multi-modal transportation networks. In sum, embracing the synergy between AI and OR is not just a technological upgrade—it’s a strategic necessity driving the future of electric vehicle fleet management.

References

[1] Kokare, B. D., Deokar, S. A., & Kale, M. M. (2025). Hybrid AI Models for Predictive Electric Vehicle Battery Capacity Estimation and Fault Tolerance Management. SSRG International Journal of Electrical and Electronics Engineering, 12(4). [2] Liu, H., et al. (2025). Multi-modal framework for battery state of health evaluation in electric vehicles. Nature Communications, 16, 1234. [3] Thakur, A., Giri, D., Panda, S., & Usharani, R. (2024). Optimizing Electric Vehicle Routing: A Statistical Analysis of ACO, GA, and SA Algorithms. International Journal of Engineering Research and Technology. [4] Rao, G., et al. (2024). Optimization of electric vehicle recharge schedule and routing: An electric vehicle routing problem with time windows. Transportation Research Part C. [5] Wipro. (2019). Electric Vehicle Battery Health Forecasting using AI/ML techniques. Wipro Analytics. [6] EVIoT-PredictiveMaint Dataset. (2025). Publicly available EV fleet sensor data. Kaggle.

Copyright

Copyright © 2025 Keya Chincholkar. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.