# Farasis Industrial Pouch-Cell Cycle Life Data: 123 Large-Format Li-ion Cells from Discovery Learning (Nature 2026), BatteryLife Processed Copy

## Facts

- Provider: Jiawei Zhang, Yifei Zhang, Baozhao Yi, Yao Ren et al. (original data); BatteryLife team, HKUST Guangzhou (processed copy)
- License: CC-BY-4.0 (https://zenodo.org/api/records/17654407)
- Size: 126 files
- Stored at: Hugging Face, https://huggingface.co/datasets/Battery-Life/BatteryLife_Farasis
- Access: public, readable without an account
- DOI: https://doi.org/10.5281/zenodo.17654407
- Category: Energy
- Source: https://doi.org/10.1038/s41586-025-09951-7
- Page: https://scigantic.com/archive/96a86c12-73bb-5315-8925-eb8f631f00c8
- Keywords: Energy, benchmark, Hugging Face

## About

Cycling data for 123 industrial-grade large-format lithium-ion pouch cells from Farasis Energy, released with Zhang et al., Discovery Learning predicts battery cycle life from minimal experiments (Nature, 2026). The cells cover eight cell types with diverse material-design combinations and cycling protocols and form 37 groups of near-identical cells, which supports study of cell-to-cell variability and early cycle-life prediction. The original Zenodo record (doi:10.5281/zenodo.17654407) provides one-second raw Parquet logs of current, voltage, temperature and step information, reference performance test (RPT) capacity workbooks and a cell-to-group map with cycle-life labels; seven mission-profile or driving-protocol cells keep only their RPT segments because the cycling portions are proprietary. This Hugging Face repository, from the BatteryLife benchmark team (HKUST Guangzhou), holds 123 processed Python pickle files, one per cell named by the original Farasis cell id (about 121 GB), with SOC and corrections applied for use with the BatteryLife preprocessing and benchmarking code (github.com/Ruifeng-Tan/BatteryLife). Original data licensed CC BY 4.0; cite Zhang et al. 2026, Nature, doi:10.1038/s41586-025-09951-7.

## Load the data

```python
from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="Battery-Life/BatteryLife_Farasis", filename="Farasis_PA-b1_001_single_parquet_with_soc_and_corrections.pkl", repo_type="dataset")
print(path)
```

## Data structure

Example file: `Farasis_PA-b1_001_single_parquet_with_soc_and_corrections.pkl`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/96a86c12-73bb-5315-8925-eb8f631f00c8
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/96a86c12-73bb-5315-8925-eb8f631f00c8
- Over MCP: `get_schema_card` with id `96a86c12-73bb-5315-8925-eb8f631f00c8` at https://api.scigantic.com/mcp

## Related topics

- [Benchmark datasets](https://scigantic.com/topics/benchmark)
- [Energy datasets](https://scigantic.com/topics/energy)
