How far can routine process records predict final turbidity in industrial beer crossflow filtration? A capability-boundary study on 5000 production batches
Source check: Source checked. Not independent scientific confirmation.
Finding: Across 5,000 brewery batches, models using routine process records explained at most about 8% of held-out turbidity variance; no nonlinear model beat linear baselines and a learning curve plateaued near R²=.10. This is a specific industrial observability/capability result, bounded to the recorded features and production setting; it is not a causal intervention result.
Technology: Regressors test how routine brewery records predict measured final turbidity
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
- AI contribution
- Regressors test how routine brewery records predict measured final turbidity
- Test and result
- Regressors test how routine brewery records predict measured final turbidity; the learned models establish a bounded process-data prediction ceiling.; PDF Results §§3.1–3.3: best linear models yield R²≈.084; multi-seed and learning-curve analyses plateau around .10, showing limited information in recorded features.; Holdout/multiseed and fixed-block learning-curve analysis directly supports the capability-boundary claim; batch data from one production setting and noninterventional design limit t
- Limitation / still unknown
- External generalization, independent replication and downstream real-world impact were not established by the reviewed source.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://www.researchsquare.com/article/rs-11487956/v1
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.