An integrated fuzzy logic and machine learning platform for porosity detection using optical tomography imaging during laser powder bed fusion
Bibliographic record
Abstract
Abstract Traditional methods such as mechanical testing and x-ray computed tomography (CT), for quality assessment in laser powder-bed fusion (LPBF), a class of additive manufacturing (AM), are resource-intensive and conducted post-production. Recent advancements in in-situ monitoring, particularly using optical tomography (OT) to detect near-infrared light emissions during the process, offer an opportunity for in-situ defect detection. However, interpreting OT datasets remains challenging due to inherent process characteristics and disturbances that may obscure defect identification. This paper introduces a novel machine learning-based approach that integrates a self-organizing map, a fuzzy logic scheme, and a tailored U-Net architecture to enhance defect prediction capabilities during the LPBF process. This model not only predicts common flaws such as lack of fusion and keyhole defects through analysis of in-situ OT data, but also allows quality assurance professionals to apply their expert knowledge through customizable fuzzy rules. This capability facilitates a more nuanced and interpretable model, enhancing the likelihood of accurate defect detection. The efficacy of this system has been validated through experimental analyses across various process parameters, with results validated by subsequent CT scans, exhibiting strong performance with average model scores ranging from 0.375 to 0.819 for lack of fusion defects and from 0.391 to 0.616 for intentional keyhole defects. These findings underscore the model’s reliability and adaptability in predicting defects, highlighting its potential as a transformative tool for in-process quality assurance in AM. A notable benefit of this method is its adaptability, allowing the end-user to adjust the probability threshold for defect detection based on desired quality requirements and custom fuzzy rules.
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 | 0.001 |
| 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 | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
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".