{"success":true,"data":{"author":"Matthias Luh, Thomas Blank (Karlsruhe Institute of Technology)","bucketAccessType":"public","bucketName":"bsebench-org/lg-hg2-stroebl-2024-raw","bucketPrefix":"","category":"Energy","cloudProvider":"huggingface","createdAt":"2026-11-19T15:09:18.457Z","description":"A subset of the Karlsruhe Institute of Technology (KIT) comprehensive battery aging dataset of Luh and Blank (Scientific Data 11:1004, 2024), which aged 228 commercial LG INR18650HG2 cells (NMC cathode, graphite with silicon oxide anode, C+SiO) for more than a year under calendar, cyclic and driving-profile conditions, over 3 billion data points at two-second resolution. This copy holds only profile-aging cells driven with the Worldwide harmonized Light-duty Test Cycle (WLTC, Class 3b passenger car, WLTP): 28 truncated measurement-log CSVs named cell_logext_P0xx_r_Sxx_Cxx_truncated.csv for parameter sets P065 to P076 (about 8.5 GB), with voltage, current, temperature and related columns. BSEBench uses it to test state of charge (SOC) estimation filters such as Kalman filters on silicon-containing anodes. The full 69.4 GB raw release, result data (capacity and impedance fade) and an updated log release are on KIT RADAR4KIT under DOIs 10.35097/1947, 10.35097/1969 and 10.35097/kww7jv8ajuvchcah; reader scripts are on GitHub (energystatusdata/bat-age-data-scripts). Note: the Hugging Face card attributes this dataset to Stroebl et al.; the KIT record and paper name Matthias Luh and Thomas Blank. Licensed CC BY 4.0; cite Luh and Blank 2024, doi:10.1038/s41597-024-03831-x.","displayTitle":"KIT Comprehensive Battery Aging Dataset, WLTC Driving-Cycle Subset: LG INR18650HG2 NMC/C-SiO Cell Logs (Luh and Blank, bsebench copy)","doi":"10.35097/1947","huggingFaceUrl":"https://huggingface.co/datasets/bsebench-org/lg-hg2-stroebl-2024-raw","id":"5b8e5697-3db4-58ba-b551-def66d6ee6e0","license":"CC-BY-4.0","licenseEvidenceUrl":"https://api.datacite.org/dois/10.35097/1947","objectCountEstimate":30,"probeStatus":"ok","referenceUrl":"https://doi.org/10.1038/s41597-024-03831-x","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"bsebench-org/lg-hg2-stroebl-2024-raw\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"bsebench-org/lg-hg2-stroebl-2024-raw\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=bsebench-org, license:cc-by-4.0, size_categories:1B<n<10B, downloads=376"],"format":"mixed","generatedAt":"2026-11-19T15:21:43.995Z","layout":{"example":"cell_logext_P065_1_S01_C03_truncated.csv","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"\"timestamp_s\";\"timestamp_origin\";\"sd_block_id\";\"v_raw_V\";\"i_raw_A\";\"p_raw_W\";\"t_cell_degC\";\"delta_q_Ah\";\"delta_e_Wh\";\"soc_est\";\"ocv_est_V\";\"meas_condition\";\"bms_condition\";\"v_state\";\"i_state\";\"t_state\";\"over_q\";\"under_q\";\"over_soc\";\"under_soc\";\"eol_cap\";\"eol_imp\";\"scheduler_state_dec\";\"scheduler_state_phase\";\"scheduler_state_checkup\";\"scheduler_state_charge\";\"scheduler_timeout\";\"scheduler_cu_pending\";\"scheduler_pause_pending\";\"scheduler_state_sub\";\"delta_q_chg_Ah\";\"delta_q_dischg_Ah\";\"delta_e_chg_Wh\";\"delta","encoding":"utf8","key":"cell_logext_P065_1_S01_C03_truncated.csv","truncatedAt":512}],"sampleObjects":["cell_logext_P065_1_S01_C03_truncated.csv","cell_logext_P065_2_S03_C02_truncated.csv","cell_logext_P065_3_S04_C02_truncated.csv","cell_logext_P066_1_S02_C03_truncated.csv","cell_logext_P066_2_S03_C03_truncated.csv"],"sidecars":[{"bytes":"# LG HG2 WLTC truncated bundle (recovered)\n# Generated: 2026-04-26T00:34:09Z\n# Source: cell_logext.7z extracted on Hetzner main 178.105.10.218\n# Truncation: first 1000000 lines per CSV\n# Recovery context: original extract finished 7z phase but the\n#   truncate step was not reached due to disk-full at peak\n#   (extracted=130GB + tar=65GB + 7z=56GB = 251GB > 246GB volume).\n#   Volume freed by removing tar + 7z (no longer needed).\n\n## Files in this bundle:\ntruncated/cell_logext_P065_1_S01_C03_truncated.csv 318462004 bytes\ntruncated/cell_logext_P065_2_S03_C02_truncated.csv 306672954 bytes\ntruncated/cell_logext_P065_3_S04_C02_truncated.csv 316930673 bytes\ntruncated/cell_logext_P066_1_S02_C03_truncated.csv 317218772 bytes\ntruncated/cell_logext_P066_2_S03_C03_truncated.csv 318492427 bytes\ntruncated/cell_logext_P066_3_S04_C03_truncated.csv 313801144 bytes\ntruncated/cell_logext_P067_1_S01_C02_truncated.csv 304451527 bytes\ntruncated/cell_logext_P067_2_S02_C02_truncated.csv 307776985 bytes\ntruncated/cell_logext_P067_3_S05_C01_truncated.csv 307923673 bytes\ntruncated/cell_logext_P068_1_S06_C03_truncated.csv 319910027 bytes\ntruncated/cell_logext_P068_2_S07_C03_truncated.csv 317143682 bytes\ntruncated/cell_logext_P068_3_S08_C03_truncated.csv 317701533 bytes\ntruncated/cell_logext_P069_1_S08_C02_truncated.csv 315963677 bytes\ntruncated/cell_logext_P069_2_S09_C02_truncated.csv 318033626 bytes\ntruncated/cell_logext_P069_3_S10_C01_truncated.csv 317968130 bytes\ntruncated/cell_logext_P070_1_S06_C02_truncated.csv 308064712 bytes\ntruncated/cell_logext_P070_2_S07_C02_truncated.csv 307644313 bytes\ntruncated/cell_logext_P070_3_S09_C01_truncated.csv 303090869 bytes\ntruncated/cell_logext_P071_1_S11_C03_truncated.csv 317844197 bytes\ntruncated/cell_logext_P071_2_S12_C03_truncated.csv 317378714 bytes\ntruncated/cell_logext_P071_3_S13_C03_truncated.csv 316730168 bytes\ntruncated/cell_logext_P072_1_S10_C07_truncated.csv 316834054 bytes\ntruncated/cell_logext_P072_2_S13_C02_truncated.csv 317558432 bytes\ntruncated/cell_logext_P072_3_S14_C02_truncated.csv 319417464 bytes\ntruncated/cell_logext_P073_1_S11_C02_truncated.csv 303686334 bytes\ntruncated/cell_logext_P073_2_S12_C02_truncated.csv 306216071 bytes\ntruncated/cell_logext_P073_3_S14_C01_truncated.csv 304920627 bytes\n\n## Original 7z entries (full sizes from wltc_files.txt):\n2024-02-22 18:32:25 ....A   5207858593               cell_logext_P065_1_S01_C03.csv\n2024-02-22 18:33:38 ....A   5931604806               cell_logext_P065_2_S03_C02.csv\n2024-02-22 18:34:49 ....A   5500184645               cell_logext_P065_3_S04_C02.csv\n2024-02-22 18:36:16 ....A   6666157757               cell_logext_P066_1_S02_C03.csv\n2024-02-22 18:37:23 ....A   6677801936               cell_logext_P066_2_S03_C03.csv\n2024-02-22 18:39:11 ....A   6654456738               cell_logext_P066_3_S04_C03.csv\n2024-02-22 18:38:05 ....A   1017434607               cell_logext_P067_1_S01_C02.csv\n2024-02-22 18:38:47 ....A   1056855906               cell_logext_P067_2_S02_C02.csv\n2024-02-22 18:39:33 ....A   1012600593               cell_logext_P067_3_S05_C01.csv\n2024-02-22 18:41:54 ....A   6693510243               cell_logext_P068_1_S06_C03.csv\n2024-02-22 18:42:05 ....A   6699908874               cell_logext_P068_2_S07_C03.csv\n2024-02-22 18:44:44 ....A   6691494455               cell_logext_P068_3_S08_C03.csv\n2024-02-22 18:44:43 ....A   6698227455               cell_logext_P069_1_S08_C02.csv\n2024-02-22 18:47:26 ....A   6689215792               cell_logext_P069_2_S09_C02.csv\n2024-02-22 18:47:29 ....A   6667121638               cell_logext_P069_3_S10_C01.csv\n2024-02-22 18:48:28 ....A   1351493344               cell_logext_P070_1_S06_C02.csv\n2024-02-22 18:48:30 ....A   1303621768               cell_logext_P070_2_S07_C02.csv\n2024-02-22 18:49:41 ....A   1880588020               cell_logext_P070_3_S09_C01.csv\n2024-02-22 18:51:15 ....A   6705360299               cell_logext_P071_1_S11_C03.csv\n2024-02-22 18:52:14 ....A   6704473954               cell_logext_P071_2_S12_C03.csv\n2024-02-22 18:53:51 ....A   6708120722               cell_logext_P071_3_S13_C03.csv\n2024-02","encoding":"utf8","key":"MANIFEST.txt","name":"MANIFEST.txt","size":5349,"truncated":true},{"bytes":"---\nlicense: cc-by-4.0\nlanguage:\n  - en\nsize_categories:\n  - 1B<n<10B\ntask_categories:\n  - time-series-forecasting\n  - tabular-regression\ntags:\n  - battery\n  - state-of-charge\n  - state-of-health\n  - lithium-ion\n  - NMC\n  - silicon-graphite\n  - LG-INR18650HG2\n  - WLTC\n  - drive-cycle\n  - dynamic-load\n  - aging\n  - kalman-filter\n  - benchmark\n  - bsebench\n  - tier-1-raw\npretty_name: \"LG INR18650HG2 Stroebl 2024 — Tier 1 Raw Mirror (truncated WLTC)\"\n---\n\n# lg-hg2-stroebl-2024-raw\n\n**Tier 1 raw mirror** of the LG INR18650HG2 cycling dataset by Stroebl, Mayer, Lienkamp (KIT, 2024), hosted under the [BSEBench](https://bsebench.org) organization.\n\nThis dataset closes a major BSEBench V1.0 coverage gap : authentic WLTC dynamic driving telemetry on NMC cells with **silicon-carbide composite anode** (C-SiO blended graphite, 3-6% SiO per Obrovac 2014). Si-C anodes exhibit path-dependent hysteresis ~10% at mid-SOC that classical 2-RC ECM cannot correct — a stress-test for ensemble-based filtering approaches.\n\n## Status\n\n🚧 **Upload pending** — placeholder card. This Tier 1 mirror contains the **truncated WLTC subset** (28 CSV files, ~7.9 GB) used in the BSEBench paper2b validation work. The full 69.4 GB raw distribution is preserved under canonical DOI [10.35097/1947](https://doi.org/10.35097/1947) ; this subset is sufficient for BSEBench V1.1 evaluation.\n\n## What this is\n\nLG Chem INR18650HG2 cells cycled with WLTC (Worldwide Harmonized Light-duty Vehicles Test Cycle) profiles across 12 parameter sets (P065-P076) × 3 replicates. Schema is plain CSV (pyarrow-readable). 5 dataset types : LOG (59 cols, 2 s resolution), EOC (57 cols per half-cycle), LOG_AGE (11 cols), EIS (22 cols), PULSE (19 cols).\n\nCitation (verbatim from Scientific Data paper) :\n\n> **Stroebl, F., Mayer, A. and Lienkamp, M. (2024).** *Open Source Cyclic Aging Dataset of 228 Commercial Lithium-Ion Battery Cells.* Scientific Data, 11, 944.\n\n## Cell specifications\n\n| Property | Value |\n|---|---|\n| Manufacturer | LG Chem |\n| Model | INR18650HG2 |\n| Form factor | 18650 cylindrical |\n| Cathode chemistry | NMC (typical 622 stoichiometry) |\n| Anode chemistry | **C-SiO composite** (3-6% SiO blended with graphite) |\n| Nominal capacity | 3.0 Ah |\n| Total cells in full dataset | 228 |\n| Cells in this truncated subset | 28 (P065-P076 × 3 reps, ~36 minus a few missing) |\n| Temperatures | 0, 10, 25, 40 °C (read from `pool_meas_t_degC` column) |\n| Aging duration | up to 449 days |\n| Cell ID format | `lg_P{param:03d}_rep{r}_S{stage:02d}` |\n\n## File inventory\n\nThis subset contains 28 CSVs named `cell_logext_P{param}_{rep}_S{stage}_C{cell}_truncated.csv` covering the 12 WLTC parameter sets with 3 replicates each.\n\n## Profile\n\nWLTC driving (Worldwide Harmonized Light-duty Vehicles Test Cycle), repeated until SoC < 10%. Sign convention in raw : `i_raw_A` positive = charge ; BSEBench Tier 2 will flip to BPX-1.1 (charge=positive) — wait, BPX-1.1 IS charge=positive so no flip needed. Verify empirically at adapter time.\n\n## Si-C anode anomaly (important caveat)\n\nPer [Obrovac 2014](https://doi.org/10.1149/2.005403jes) and [McDowell 2013](https://doi.org/10.1002/adma.201301795) : Si-containing anodes produce path-dependent hysteresis ~10% at mid-SOC, **uncorrectable by classical 2-RC ECM**. This dataset is therefore a stress-test for advanced filter approaches (BMA ensembles, neural state-space, particle-filter mixture models). BSEBench filter benchmarks should **explicitly report Si-C-vs-graphite performance differential**.\n\n## Why \"raw mirror\" tier (truncated)\n\nBSEBench dual-tier strategy :\n- **Tier 1 (this repository)** — truncated WLTC subset (28 CSVs, 7.9 GB), SHA-256 verified\n- **Tier 2** — see [`bsebench-org/lg-hg2-stroebl-2024`](https://huggingface.co/datasets/bsebench-org/lg-hg2-stroebl-2024) for harmonized BSEBench-canonical Parquet\n- **Full raw** : the canonical 69.4 GB distribution under DOI [10.35097/1947](https://doi.org/10.35097/1947) is **NOT mirrored** in BSEBench V1.1 — interested users should download directly from the KIT data portal. BSEBench mirror cove","encoding":"utf8","key":"README.md","name":"README.md","size":5625,"truncated":true}],"sourceObjectKey":"cell_logext_P065_1_S01_C03_truncated.csv","starterCell":"import pandas as pd\ndf = pd.read_csv(\"hf://datasets/bsebench-org/lg-hg2-stroebl-2024-raw/cell_logext_P065_1_S01_C03_truncated.csv\")\ndf.head()","starterSamplePath":"cell_logext_P065_1_S01_C03_truncated.csv","totalObjects":30},"summary":"A subset of the Karlsruhe Institute of Technology (KIT) comprehensive battery aging dataset of Luh and Blank (Scientific Data 11:1004, 2024), which aged 228 commercial LG…","title":"KIT Comprehensive Battery Aging Dataset, WLTC Driving-Cycle Subset: LG INR18650HG2 NMC/C-SiO Cell Logs (Luh and Blank, bsebench copy)","updatedAt":"2026-11-19T19:10:17.419Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/bsebench-org/lg-hg2-stroebl-2024-raw"}}