celljar

Public battery cell test data, harmonized and sealed in one schema (Parquet + JSON).

celljar reads raw files from published sources and writes them into one canonical schema across four entities: cell_metadata + test_metadata (JSON), timeseries + cycle_summary (Parquet). Consumers read one format instead of writing per-source loaders.

Scope: harmonization only. celljar focuses on measurements - unit conversion and schema normalization. It deliberately leaves fitting and modeling to downstream tools that specialize in those steps.

• Upstream code / issue tracker: https://github.com/mihnathul/celljar
• Sources in this snapshot: BILLS, CLO, ECKER, HNEI, MATR, NASA_PCOE, NAUMANN, ORNL
• Contents: 273 cells, 348 tests, 167,820,250 timeseries rows

Files

cells/*.json              # one file per cell (hardware metadata)
tests/*.json              # one file per test (protocol + provenance + observed stats)
timeseries.parquet        # all tests' V/I/T samples; join on test_id
cycle_summary.parquet     # per-cycle aggregates (aging studies); join on (test_id, cycle_number)

Schema (overview)

Four entities; field list generated from the authoritative JSON Schemas:

• cell_metadata (JSON, one file per cell) - cell_id, source, source_cell_id, manufacturer, model_number, chemistry, cathode, anode, electrolyte, form_factor, nominal_capacity_Ah, nominal_voltage_V, max_voltage_V, min_voltage_V
• test_metadata (JSON, one file per test) - test_id, cell_id, test_type*, temperature_C_min, temperature_C_max, soc_range_min, soc_range_max, soc_step, c_rate_charge, c_rate_discharge, protocol_description, num_cycles, soh_pct, soh_method, cycle_count_at_test, test_year, source_doi, source_url, source_citation, source_license, source_license_url, n_samples, duration_s, voltage_observed_min_V, voltage_observed_max_V, current_observed_min_A, current_observed_max_A, temperature_observed_min_C, temperature_observed_max_C, sample_dt_min_s, sample_dt_median_s, sample_dt_max_s
• timeseries (Parquet, one row per measurement sample) - test_id, cycle_number, step_number, step_type, timestamp_s*, voltage_V, current_A, temperature_C, coulomb_count_Ah, energy_Wh, displacement_um
• cycle_summary (Parquet, one row per cycle / aging checkpoint) - test_id, cell_id, cycle_number, equivalent_full_cycles, elapsed_time_s, capacity_Ah, capacity_retention_pct, resistance_dc_ohm, resistance_dc_pulse_duration_s, resistance_dc_soc_pct, energy_Wh, coulombic_efficiency, temperature_C_mean

* = required field (others nullable). See JSON Schemas for full type info, enum values, and constraints.

SI units. Relative timestamps. Missing data is explicit null (no NaN sentinels). Current sign convention: positive = charge (into the cell), negative = discharge.

Join keys: cells.cell_id = tests.cell_id, tests.test_id = timeseries.test_id, (tests.test_id, cycle_number) = cycle_summary.(test_id, cycle_number).

Download the whole bundle

# CLI - pulls everything (cells/*.json, tests/*.json, timeseries.parquet, cycle_summary.parquet)
pip install huggingface_hub
huggingface-cli download mihnathul/celljar --repo-type dataset --local-dir ./celljar-bundle

# Pin a tagged release for reproducibility
huggingface-cli download mihnathul/celljar --repo-type dataset --revision v0.2.0 --local-dir ./celljar-bundle

Or in Python:

from huggingface_hub import snapshot_download
local = snapshot_download(repo_id="mihnathul/celljar", repo_type="dataset", revision="v0.2.0")
print(local)  # local path containing cells/, tests/, timeseries.parquet, cycle_summary.parquet

Query in place - no download needed

DuckDB - full SQL across all entities over HTTPS

INSTALL httpfs; LOAD httpfs;
SELECT c.chemistry, c.nominal_capacity_Ah,
       t.test_id, t.test_type, t.soh_pct,
       COUNT(*) AS n_samples
FROM read_json('https://huggingface.co/datasets/mihnathul/celljar/resolve/main/cells/*.json')  c
JOIN read_json('https://huggingface.co/datasets/mihnathul/celljar/resolve/main/tests/*.json')  t
  ON c.cell_id = t.cell_id
JOIN 'https://huggingface.co/datasets/mihnathul/celljar/resolve/main/timeseries.parquet'       ts
  ON t.test_id = ts.test_id
GROUP BY 1,2,3,4,5
ORDER BY t.test_id;

pandas / Polars - predicate-pushdown read of one test

import pandas as pd
df = pd.read_parquet(
    "https://huggingface.co/datasets/mihnathul/celljar/resolve/main/timeseries.parquet",
    filters=[("test_id", "==", "ORNL_LEAF_2013_HPPC_25C")],
)

datasets library - streaming

from datasets import load_dataset
ds = load_dataset(
    "parquet",
    data_files="https://huggingface.co/datasets/mihnathul/celljar/resolve/main/timeseries.parquet",
    split="train",
    streaming=True,
)
for row in ds.take(5):
    print(row)

License & citation

The science here belongs to the original authors; celljar simply puts their data in one place with a shared schema. Please cite their papers when you use the data, and, if it's helpful, celljar alongside.

• This harmonized bundle (packaging, schema, derived test-metadata fields): CC-BY-4.0.
• Upstream raw data retains each publisher's original license - listed per-source below. Each source's license terms apply when you use its tests.

To make attribution easy, every tests/*.json row carries its own source_doi, source_citation, source_license, and source_license_url fields, so you can pull references for any analysis with one query.

Per-source citations

BILLS

Bills, A., Sripad, S., Fredericks, W. L., et al. (2023). A battery dataset for electric vertical takeoff and landing aircraft. Scientific Data 10, 344. https://doi.org/10.1038/s41597-023-02180-5

License: CC-BY-4.0 · license terms · dataset · DOI: 10.1184/R1/14226830

CLO

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

License: CC-BY-4.0 · license terms · dataset · DOI: 10.1038/s41586-020-1994-5

ECKER

(citation unavailable in harmonized bundle)

License: see upstream

HNEI

Kollmeyer, P. (2018). Panasonic 18650PF Li-ion Battery Data. Mendeley Data, v1. https://doi.org/10.17632/wykht8y7tg.1

License: CC-BY-4.0 · license terms · dataset · DOI: 10.17632/wykht8y7tg.1

MATR

Severson, K. A., Attia, P. M., Jin, N., et al. (2019). Data-driven prediction of battery cycle life before capacity degradation. Nature Energy 4, 383-391. https://doi.org/10.1038/s41560-019-0356-8

License: CC-BY-4.0 · license terms · dataset · DOI: 10.1038/s41560-019-0356-8

NASA_PCOE

Saha, B. & Goebel, K. (2007). Battery Data Set. NASA Prognostics Data Repository, NASA Ames Research Center, Moffett Field, CA. https://www.nasa.gov/intelligent-systems-division/discovery-and-systems-health/pcoe/pcoe-data-set-repository/ Cells are 18650 Li-ion; chemistry/vendor not disclosed by NASA — community consensus treats them as LCO.

License: CC0-1.0 · license terms · dataset

NAUMANN

Naumann, M. (2021). Data for: Analysis and modeling of calendar/cycle aging of a commercial LiFePO4/graphite cell. Mendeley Data. DOIs: 10.17632/kxh42bfgtj.1 (calendar) and 10.17632/6hgyr25h8d.1 (cycle). Companion papers: Naumann et al. JPS 2018 doi:10.1016/j.est.2018.01.019, Naumann et al. JPS 2020 doi:10.1016/j.jpowsour.2019.227666

License: CC-BY-4.0 · license terms · dataset · DOI: 10.17632/kxh42bfgtj.1

ORNL

Wiggins, G., Allu, S., & Wang, H. (2019). Battery cell data from a 2013 Nissan Leaf. Oak Ridge National Laboratory. https://doi.org/10.5281/zenodo.2580327

License: MIT · license terms · dataset · DOI: 10.5281/zenodo.2580327

Citing celljar

If you'd like to cite celljar:

@software{celljar,
  author = {Mihna Neerulpan},
  title  = {celljar: Public Battery Test Dataset Harmonization with a Canonical Schema},
  year   = {2026},
  url    = {https://github.com/mihnathul/celljar},
}

Links

• Code: https://github.com/mihnathul/celljar
• Issues / new-source requests: https://github.com/mihnathul/celljar/issues
• Canonical JSON Schemas: https://github.com/mihnathul/celljar/tree/main/schemas