Identification and Characterization of an Electric Vehicle's Operational Regimes Using Sensor Data

The increasing adoption of electric vehicles (EVs) has led to the generation of large-scale multi-domain datasets containing information on battery, thermal, charging, and driving behavior under real-world operating conditions. These datasets provide valuable insights into vehicle operation by enabling the analysis of interactions among electrical, environmental, and operational variables. This paper presents a comprehensive analysis of a real-world EV dataset collected during on-road operation, focusing on the characterization of the available measurements and their suitability for describing different vehicle operating conditions. The dataset is analyzed through statistical evaluation, correlation analysis, and dimensionality reduction techniques to investigate variable relationships and the representation of vehicle behavior. Specific operating conditions, including driving, charging, regenerative braking, and parking states, are described based on the available sensor measurements and reconstructed information where required. The results highlight the potential and limitations of real-world EV datasets, demonstrating the importance of data quality, variable consistency, and operational state reconstruction for accurately interpreting vehicle behavior and supporting future data-driven mobility applications.