Synthetic Data-Driven Mixed Reality for AR-Assisted Maintenance
Bibliographic record
Abstract
AI-assisted augmented reality (AR) tools hold significant promise for improving industrial maintenance and assembly workflows, particularly for training novice technicians, amid a shortage of experienced personnel in the Aerospace sector. This work introduces an end-to-end methodology for developing AI-assisted AR tools only using 3D CAD data. The proposed workflow is especially valuable to industries-like Aerospace-that manage comprehensive 3D data. The proposed pipeline uses a context-oriented training strategy using fully assembled CAD models to generate domain-randomized datasets for component recognition. At runtime, assembly pose data is used to project regions of interest onto the image. The trained component recognition model verifies if the component is installed within each projected region. To validate the data generation methodology, a sample gearbox containing four unique components was modeled to produce a synthetic dataset of 13,000 images, which was used to train Faster R-CNN for component recognition. The model was evaluated on a training set consisting of 112 real images of the gearbox. After 30 epochs of training, the component recognition model-evaluated on the test set across different IoU thresholds (0.50-0.95) and confidence scores (0.05-0.95)-demonstrated state of the art precision and recall results based on dataset composition. This work demonstrates that AI-assisted AR tools can be realized entirely from synthetic data produced through the proposed pipeline, removing the traditional bottleneck of real-world image collection and manual annotation.
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 | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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".