Learning More with Less: A Generalizable, Self-Supervised Framework for Privacy-Preserving Capacity Estimation with EV Charging Data

TL;DR

This work tackles the challenge of estimating EV battery capacity under strict privacy and labeling constraints. It introduces a self-supervised pre-training framework trained on privacy-preserving charging snippets, employing snippet similarity-weighted masked reconstruction and snippet-wise contrastive learning to learn rich, transferable representations, followed by supervised fine-tuning on a small labeled set. Empirical results demonstrate state-of-the-art generalization under age- and manufacturer-induced distribution shifts, achieving up to $31.9\%$ lower RMSE than baselines while using only $10\%$ of labeled data. The approach offers a practical, privacy-preserving path to data-efficient prognostics and can be extended to other sensor-rich domains where data fragmentation and privacy concerns are prominent.

Abstract

Accurate battery capacity estimation is key to alleviating consumer concerns about battery performance and reliability of electric vehicles (EVs). However, practical data limitations imposed by stringent privacy regulations and labeled data shortages hamper the development of generalizable capacity estimation models that remain robust to real-world data distribution shifts. While self-supervised learning can leverage unlabeled data, existing techniques are not particularly designed to learn effectively from challenging field data -- let alone from privacy-friendly data, which are often less feature-rich and noisier. In this work, we propose a first-of-its-kind capacity estimation model based on self-supervised pre-training, developed on a large-scale dataset of privacy-friendly charging data snippets from real-world EV operations. Our pre-training framework, snippet similarity-weighted masked input reconstruction, is designed to learn rich, generalizable representations even from less feature-rich and fragmented privacy-friendly data. Our key innovation lies in harnessing contrastive learning to first capture high-level similarities among fragmented snippets that otherwise lack meaningful context. With our snippet-wise contrastive learning and subsequent similarity-weighted masked reconstruction, we are able to learn rich representations of both granular charging patterns within individual snippets and high-level associative relationships across different snippets. Bolstered by this rich representation learning, our model consistently outperforms state-of-the-art baselines, achieving 31.9% lower test error than the best-performing benchmark, even under challenging domain-shifted settings affected by both manufacturer and age-induced distribution shifts. Source code is available at https://github.com/en-research/GenEVBattery.

Figures (8)

• Figure 1: Contribution of our work to the domain of practical EV battery capacity estimation using field data.
• Figure 2: Visualization of the three generated data distributions. Considering the charging snippets across all EVs, (a) illustrates the inverse relationship between the mileage and capacity labels of the charging snippets, particularly, highlighting the capacity degradation that occurs through distributions 1 to 3. (b) This capacity degradation results in the notable domain shifts.
• Figure 3: (a) An overview of "pre-training--fine-tuning" pipeline for developing a generalizable capacity estimation model, followed by (b) an in-depth look at our proposed pre-training framework. The key innovation is learning meaningful similarities among the fragmented snippets, which enables the model to focus on similarities between the masked and original snippets, while suppressing noise from dissimilar snippets for the masked reconstruction process. With the snippet-wise contrastive learning and subsequent similarity-weighted masked reconstruction, the model captures rich representations of both granular charging patterns within individual snippets and high-level associative relationships across different snippets, significantly enhancing downstream capacity estimation.
• Figure 4: Comparison of some randomly sampled current patterns from the novel test data with typical patterns observed in labeled and unlabeled snippets.
• Figure 5: Inference performance on novel test patterns beyond fine-tuning data.