---
title: Closed-loop optimization of extreme fast charging for batteries
license: cc-by-4.0
pretty_name: Closed-loop optimization of extreme fast charging for batteries
tags:
- bsebench
- batteries
- fast-charging
- raw-mirror
- materials-science
---

# Closed-loop optimization of extreme fast charging for batteries

BSEBench status: `raw_mirror_pending_validation`

This dataset repository is a raw mirror of the TRI Energy & Materials Data / data.matr.io project "Closed-loop optimization of extreme fast charging for batteries using machine learning". The source describes commercial lithium-ion phosphate (LFP)/graphite cells cycled under fast-charging conditions for the associated Nature publication.

Source page: https://data.matr.io/  
Project URL: https://data.matr.io/1/projects/5d80e633f405260001c0b60a  
DOI: https://doi.org/10.1038/s41586-020-1994-5  
License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)  
Source institution/credibility: Stanford University / MIT / Toyota Research Institute  
Download date: 2026-05-08

## Contents

The files under `source_files/` are the official source downloads retrieved from `https://data.matr.io/1/api/v1/file/<file_id>/download` and organized by source file kind and batch. No scientific data values were transformed.

Actual downloaded source payload:

| Kind | Files | Bytes |
|---|---:|---:|
| MATLAB struct files | 5 | 3975095856 |
| JSON structure ZIP files | 5 | 6855320161 |
| Raw CSV data files | 236 | 11275069907 |
| Metadata CSV files | 236 | 124514 |
| Schedule files | 236 | 5038701 |
| Total source files | 718 | 22110649139 |

`SOURCE_FILE_MANIFEST.json` records the official file IDs, source URLs, source API metadata, local paths, and measured local byte counts. `SHA256SUMS.txt` records checksums for uploaded files.

## Provenance Notes

The TRI catalog page lists this dataset with license CC BY 4.0, and the project API page includes the citation and DOI for Attia et al. (2020). The Nature article data-availability section links the study datasets to `https://data.matr.io/1/`.

Some `size` values returned by the source API differ from the byte counts of the redirected downloadable assets, especially JSON-structure ZIP archives. This mirror treats the actual downloaded files and SHA-256 checksums as authoritative and records both API-reported and measured sizes in `SOURCE_FILE_MANIFEST.json`.

## Citation

Attia, P. M., Grover, A., Jin, N. et al. Closed-loop optimization of fast-charging protocols for batteries with machine learning. Nature 578, 397-402 (2020). https://doi.org/10.1038/s41586-020-1994-5

## License

The source page and API configuration identify the dataset license as Creative Commons Attribution 4.0 International (CC BY 4.0). Users should follow the source license and cite the original work.