{"paper":{"title":"Causal Anomaly Detection for Lithium-Ion Battery Degradation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Causal anomaly detection on routine battery measurements identifies degradation up to 402 cycles before standard failure indicators.","cross_cats":["cond-mat.stat-mech","cs.LG","physics.comp-ph"],"primary_cat":"cond-mat.mtrl-sci","authors_text":"Dieter W. Heermann, Hagen Heermann","submitted_at":"2026-05-17T08:53:16Z","abstract_excerpt":"Reliable early detection of lithium-ion battery degradation requires health indicators that are physically interpretable and computable from routine cycler telemetry without access to the degradation region. We introduce \\textsc{CausalHealth}, a framework that applies causal graph discovery and $k$-nearest-neighbour transfer entropy to per-cycle voltage, current, temperature, and resistance time series, and organises twelve resulting anomaly scores into three signal-class bundles (Magnitude-shift, Predictive-residual, Complexity-entropy) -- with Isolation Forest reported separately as it falls"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"The Magnitude-shift class achieves 100% detection across all seven tested cells spanning LFP (MIT–Stanford MATR) and LCO (NASA PCoE, CALCE CS2) chemistries, with a lead time of up to 402 cycles before conventional capacity-threshold failure on gradual-fade cells.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that the twelve anomaly scores derived from causal graph discovery and k-nearest-neighbour transfer entropy on per-cycle time series reliably reflect physical degradation processes (rather than dataset-specific artifacts), which is taken to be supported by the EIS correlation on one additional NMC cell.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"CausalHealth detects lithium-ion battery degradation with 100% sensitivity and up to 402-cycle lead time using causal anomaly scores from voltage, current, temperature, and resistance time series across seven cells.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Causal anomaly detection on routine battery measurements identifies degradation up to 402 cycles before standard failure indicators.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"8a9822b7679d5d93b6f219f08eca9e24f42216a32972c5ef816cfa69abfd1925"},"source":{"id":"2605.17334","kind":"arxiv","version":1},"verdict":{"id":"e30ddb61-3186-478f-aa72-b887d7b19973","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T22:43:39.764805Z","strongest_claim":"The Magnitude-shift class achieves 100% detection across all seven tested cells spanning LFP (MIT–Stanford MATR) and LCO (NASA PCoE, CALCE CS2) chemistries, with a lead time of up to 402 cycles before conventional capacity-threshold failure on gradual-fade cells.","one_line_summary":"CausalHealth detects lithium-ion battery degradation with 100% sensitivity and up to 402-cycle lead time using causal anomaly scores from voltage, current, temperature, and resistance time series across seven cells.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that the twelve anomaly scores derived from causal graph discovery and k-nearest-neighbour transfer entropy on per-cycle time series reliably reflect physical degradation processes (rather than dataset-specific artifacts), which is taken to be supported by the EIS correlation on one additional NMC cell.","pith_extraction_headline":"Causal anomaly detection on routine battery measurements identifies degradation up to 402 cycles before standard failure indicators."},"integrity":{"clean":false,"summary":{"advisory":0,"critical":1,"by_detector":{"doi_compliance":{"total":1,"advisory":0,"critical":1,"informational":0}},"informational":0},"endpoint":"/pith/2605.17334/integrity.json","findings":[{"note":"Identifier '10.1021/je034259d' is syntactically valid but the DOI registry (doi.org) returned 404, and Crossref / OpenAlex / internal corpus also have no record. The cited work could not be located through any authoritative source.","detector":"doi_compliance","severity":"critical","ref_index":4,"audited_at":"2026-05-19T22:52:32.844378Z","detected_doi":"10.1021/je034259d","finding_type":"unresolvable_identifier","verdict_class":"cross_source","detected_arxiv_id":null}],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T23:01:19.678521Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T22:52:32.844378Z","status":"completed","version":"1.0.0","findings_count":1},{"name":"claim_evidence","ran_at":"2026-05-19T21:41:57.807085Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.741443Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"8aa8945913ea3e287d8a7c5664f63c937a2d9a6b106e603923ba1f8d64d2d297"},"references":{"count":23,"sample":[{"doi":"10.1038/s41586-020-1994-5","year":2020,"title":"Closed- loop optimization of fast-charging protocols for batteries with machine learning","work_id":"6e61a314-10c5-4413-9562-f27d2cd3984c","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1016/j.jpowsour.2016.12.011","year":2017,"title":"Degradation diagnostics for lithium ion cells","work_id":"43528639-ee47-45c5-81f7-194a9e14bf0b","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.3115/v1/d14-1179","year":2014,"title":"Learning Phrase Representations using","work_id":"938a9ac5-57f8-472b-baa1-4fbb755371fb","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1021/je034259d","year":2004,"title":"Electrolyticconductivityandglasstransitiontemperature as functions of salt content, solvent composition, and temperature for LiPF6 in propylene carbonate–ethylene carbonate","work_id":"8d8e21a5-9fe4-4e7a-9d98-4cefde483504","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.3389/fenrg","year":2022,"title":"Exploring the emerging role of large language models in smart grid cybersecurity: A survey of attacks, detection mechanisms, and mitigation strategies","work_id":"3c0a6baf-515b-4ecb-a31c-4932dc402442","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":23,"snapshot_sha256":"0ed1615fd9ad83cff87a4c5d0649b715aec7b543d443994a8ad23e158f8a7da8","internal_anchors":1},"formal_canon":{"evidence_count":2,"snapshot_sha256":"8c50f3a772801c1c942bd23cbded0aa60880f5f8ccea7a8c6a89901ec9e989f0"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}