Scope

This dataset comprises detailed data from 199 coin cell batteries featuring either NMC//graphite or LFP//graphite chemistries cycled for 1000 cycles. All batteries were assembled and cycled using the automated robotic battery materials research platform, Aurora, at Empa, the Swiss Federal Laboratories for Materials Science and Technology, within the Laboratory for Materials for Energy Conversion.

This dataset accompanies the publication:
Toward an Autonomous Robotic Battery Materials Research Platform Powered by Automated Workflow and Ontologized Findable, Accessible, Interoperable, and Reusable Data Management.
Batteries & Supercaps, 2025, https://doi.org/10.1002/batt.202500155

Data Structure

The dataset is packaged as a RO-Crate (Research Object Crate), facilitating standardized and machine-readable data sharing.

For each battery cell, the following files are included:

Battery cell metadata file:

• Filename: empa__ccid000XXX.metadata.json

• Description: Contains semantically annotated battery assembly data along with the cycling protocol applied to the specific battery cell. The file is structured in JSON-LD format and leverages the Battery Interface Ontology (BattINFO) domain to ensure semantic interoperability and rich data description.

• Reference: Battery Interface Ontology — BattInfo documentation

Battery cell cycling data files:

• Filenames:

  • empa__ccid000XXX.bdf.csv

  • empa__ccid000XXX.bdf.parquet

• Description: Cycling data is provided in two formats: CSV and Parquet. Both contain identical data. The CSV format enables straightforward human inspection, while the Parquet format supports faster parsing and is optimized for automated processing. These files follow the Battery Data Format (BDF) for Time-Series Data, as recommended by the Battery Data Alliance.

• Reference: Battery Data Alliance - Battery Data Format Definition