Lazy Fusion of Multimodal Sensors for Cost-Effective Process Monitoring
Bibliographic record
Abstract
High Resolution Image Download MS PowerPoint Slide Advances in sensing technologies and AI have resulted in new in-line and online process measurements based on video, vibration, chromatograms, and other high-dimensional data that can complement common process measurements such as pressure, temperature, and flow rates. These sensors can be beneficial for process monitoring; however, their continuous use is often highly expensive or even impractical. In this work, we propose a novel fusion strategy to integrate insights from these sources when needed while predominantly relying on the less expensive common measurements. A hierarchical organization of sensors based on a generalized cost metric serves as the basis for the fusion. The fusion process intelligently utilizes the least expensive data first. Costlier data are used by the fusion scheme only if found necessary in real-time to improve performance. Through this lazy fusion strategy, heterogeneous multimodal sensors can be utilized within a unified framework to improve decision timeliness, accuracy, and reliability while being robust to data delays, sensor failures, and computational limitations. The proposed fusion technique has been tested on two case studies, a simulated CSTR process and an experimental data set obtained from a multiphase flow facility. The obtained results show a significant reduction in diagnostic delay compared to traditional process monitoring while utilizing costly video and high-frequency measurements only 15–30% of the time.
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 | 0.002 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".