Early detection of physical fatigue in industry using wearable sensors and contextual modeling
Bibliographic record
Abstract
Physical fatigue in repetitive production lines contributes to musculoskeletal disorders and absenteeism. This study investigates a pharmaceutical packaging environment in Colombia with 43 operators (42 female; 19–53 years) performing repetitive inspection and packing. Smartwatches captured pulse rate, electrodermal activity, skin temperature, and motion, complemented by demographic (age, experience) and occupational factors (task load, line, shift, timing). Principal Component Analysis (PCA) reduced dimensionality, and a fuzzy logic–based labeling method—adapted from prior controlled experiments—generated binary and four-class fatigue labels without mid-shift self-reports. These labeled datasets were used to train multiple machine-learning classifiers. Integrating contextual features with biometrics substantially improved performance: in binary classification, F1 increased from 0.8848 (biometrics only) to 0.9375; in four-level classification, F1 rose from 0.8232 to 0.8793. Motion-related metrics emerged as the most informative predictors. Critically, feature integration improved reliability: accuracy for intermediate states (Higher Non-Fatigue and Higher Fatigue) rose by ∼10 percentage points, while false negatives in the Pure Fatigue class were eliminated—3% of cases previously misclassified as Higher Non-Fatigue were instead correctly mapped within the fatigue spectrum. This shift strengthens the system’s effectiveness for real-time safety interventions. The novelty of this work lies in combining biometric and contextual modeling to reduce false negatives in critical fatigue states, providing a scalable, non-intrusive, and human-centered early-warning system. By aligning with Industry 5.0, this approach demonstrates how wearable and contextual data can jointly support proactive and trustworthy safety interventions while maintaining operational flow.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".