Knowledge elicitation techniques and application to nuclear plant maintenance
Bibliographic record
Abstract
The new millennium has brought with it the opportunity of global trade which in turn requires the utmost in efficiency from each individual industry. This includes the nuclear power industry, a point which was emphasized when the electrical generation industry began to be de regulated across North America the late 1990s and re-emphasized when the northeast power grid of North America collapsed in the summer of 2003. This dissertation deals with reducing the cost of the maintenance function of Candu nuclear power plants and initiating a strong link between universities and the Canadian nuclear industry. This dissertation ultimately speaks to the great opportunity being presented by this approach at this time: of capturing the baby-boom generation's huge pool of knowledge before those people retire. It is expected that the protocols and procedures referenced here will have applicability across the many disciplines where collecting expert information from a similar age group is required. Various forms of RCM (reliability-centred maintenance) have been the tools of choice in industry for improving the maintenance function during the last 20 years. In this project, pilot studies, conducted at Bruce Power between 1999 and 2005, and reported on in this dissertation, lay out a path to implement statistical improvements as the next step after RCM in reducing the cost of the maintenance. Elicitation protocols, designed for the age group being elicited, address the much-documented issue of a lack of data. Clear, graphical, inferential statistical interfaces are accentuated and developed to aid in building the teams required to implement the various methodologies and to help in achieving funding targets. Graphical analysis and Crow/AMSAA (army materials systems analysis activity) plots are developed and demonstrated from the point of view of justifying the expenditures of cost reduction efforts.
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 | 0.027 | 0.079 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.005 |
| Science and technology studies | 0.002 | 0.003 |
| Scholarly communication | 0.004 | 0.004 |
| Open science | 0.002 | 0.005 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.007 | 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".