An analysis of machine shape defects in British Columbia sawmills and their classification using neural networks
Bibliographic record
Abstract
Ideally, the edges of lumber are parallel to each other and its ends are rectangular and in line with each other. However, sub-optimal occurrences in the sawing processes cause deviations from this ideal shape. In the sawmill, these deviations are often detected as off-size variations in thickness, and one particular defect shape is not necessarily distinguished from another in the downgrading process. These defects, different from machine defects like torn grain or skip, are referred to as machine shape defects in this thesis. The first part of this thesis implements a survey to analyse machine shape defects in British Columbia sawmills, while the second part employs neural networks as an experimental approach in the classification of these defects. A survey was designed and implemented to determine the industrial significance of machine shape defects in British Columbia sawmills. Completed in 2000, the survey focused on six machine shape defects commonly caused by the sawing process: snipe, flare, wedge, taper, thin snake and fat snake. Responses came from mills located across BC and from both large and small forest companies responsible for 33% of BC softwood lumber production in 2000. Characterising BC sawmills according to machine shape defects and annual production shows that for each category of mill, with one exception, there is over a 20% probability of producing at least five types of machine shape defects. The most common grade cited for all machine shape defects was No. 2 Structural. By ranking the machine shape defects in terms of occurrence and by determining which ones are most serious in terms of final quality, it was established that thin snake, snipe and taper have the most serious impact on the industry. Neural networks were trained to detect and classify snipe in rough green lumber, using more than one hundred trim ends sampled randomly from a mill experiencing difficulty processing frozen wood. A self-contained measuring apparatus was constructed to support measuring equipment and to convey the sample boards through the measuring range of six lasers at a steady rate, using the automatic feedrollers of a shaper table. A statistical model was developed to interpret the physical characteristics of the board's surface, focussing on its shape. This model was used to preprocess the laser data into a set of variables, simplifying the data set for input into the neural networks. It was demonstrated that neural networks can be applied with limited success to detect machine shape defects, in particular snipe, in random samples of rough green lumber. However, it was established that more training data is required to train the neural networks to classify the sample cases with combination snipe.
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 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) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, 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".