Explainable Quantum AI for optimizing vehicular energy management in smart cities

In rapidly growing cities, the move toward Autonomous Electric Vehicles (AEVs) is challenging the current Energy Management Systems (EMS). The goal in smart cities is to reduce emissions and improve efficiency by optimizing vehicular energy; however, it remains challenging to address real-time decisions, complex AI, and extensive computing requirements for this task. Although AI and optimization are regularly used, they cannot be trusted in safety-related situations due to issues with complexity, scalability, and lack of clarity in their actions. To achieve transparent, smart energy systems in future transportation, it is crucial to address these issues. This research proposes an Explainable Quantum AI (XQAI) model that combines the computational capabilities of Quantum Machine Learning (QML) with the interpretability of Explainable AI (XAI). With QML, dealing with complex vehicular data is more efficient, and the model uses Local Interpretable Model-Agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP) to ensure transparency and interpretability in the model’s decision-making process. This proposed model is developed using data from real cities, encompassing a wide range of features, to predict vehicular energy consumption across various trip types accurately and to provide insight into the reasons behind these predictions. According to simulation results, the proposed XQAI model is effective, as the Hybrid Classical–Quantum Regressor shows superior prediction performance with an R 2 score of 0.8439. Furthermore, using LIME revealed a confidence score of 0.95, further establishing its credibility, interpretability, and reliability. The results demonstrate that the model meets the needs for scalable, understandable, and regulated vehicular energy forecasting in smart cities.