Dataset for : Realistic Multi-Fault Diagnostics of Millions-Scale Li-ion batteries with Rapid Unsupervised Learning

Dataset for : Realistic Multi-Fault Diagnostics of Millions-Scale Li-ion batteries with Rapid Unsupervised Learning Abstract The rapid deployment of battery swapping stations necessitates scalable and reliable fault diagnosis, yet massive, sparse operational data and scarce labeled samples make this challenging. Here, we report a rapid unsupervised learning framework for realistic multi-fault diagnosis in million-scale battery fleets. Our approach employs a double-layer mechanism. First, we rapidly screen for abnormal devices by extracting features from voltage-envelope sequences. Subsequently, we pinpoint faulty cells and types using an enhanced two-stage unsupervised clustering combined with rule-based fault tracing. The framework is validated on a production dataset of over 128,000 devices, achieving 97.33% device-layer and 99.66% cell-layer accuracy. Laboratory tests on recalled batteries further confirm the detection of low-capacity and micro-short-circuit faults. These results demonstrate scalability and robustness under sparse-data conditions, enabling reliable operations for large-scale energy storage systems. Dataset Structure DataRepo/ ├── fullDataset/ │ └── fullDataset.json # Feature data for all devices └── predefinedDataset/ ├── data/ # Raw data for predefined devices └── processedData/ ├── predefinedFeatures.json # Extracted features for predefined devices ├── device_level_info.csv # Device-level information extracted from JSON └── cell_level_info.csv # Cell-level information extracted from JSON Dataset Description 1. Predefined Dataset (predefinedDataset/) The predefined dataset contains data for a selected set of devices used in preliminary research: Raw Data ( data/ ) Contains raw voltage data files for predefined devices Each file represents voltage measurements from a single device Data format: CSV files with timestamp and voltage readings for each cell Processed Data ( processedData/ ) Contains feature-extracted data from the raw measurements predefinedFeatures.json : JSON file with device filenames as keys Each value contains comprehensive feature information for the corresponding device Features include both device-level and cell-level characteristics device_level_info.csv : CSV file extracted from the JSON Contains only device-level information for easier visualization and analysis cell_level_info.csv : CSV file extracted from the JSON Contains only cell-level information for focused analysis 2. Full Dataset (fullDataset/) The full dataset contains feature-extracted data for all devices: fullDataset.json : JSON file with the same structure as predefinedFeatures.json Contains feature data for all devices in the dataset Format and feature definitions are consistent with the predefined dataset