AI-guided bridge deck inspection using vehicle-mounted infrared imaging and ultrasound tomography
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Traditional bridge deck inspections often involve manual labour and data recording, which can be time-consuming and error-prone. Infrared (IR) has strong potential to improve inspections. However, manual IR data processing can also be time-consuming. This paper presents a methodology that integrates IR imaging, an Artificial Intelligence (AI) model and Ultrasound Tomography (UT) to streamline and enhance inspections. The approach begins with vehicle-mounted IR imaging for rapid, large-scale scanning of bridge decks to identify potential concerns. Unlike conventional methods and earlier AI-integrated studies relying on pre-processed IR data, this approach uses processed IR data to label unprocessed images, which are then used to train a Grounding DINO AI model. The trained model autonomously detects and localises suspicious regions directly from raw IR images, eliminating labour-intensive processing and enabling real-time defect detection. UT is subsequently employed to provide detailed analysis of flagged areas, offering insights into damage types such as delamination, spalling and defect depth. The AI model’s precision, around 90%, is evaluated against ground truth from processed IR data. Integrating these technologies, the proposed method offers great potential to accelerate inspections, improve defect localisation with global coordinates and support efficient maintenance planning.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it