Developing a national dataset of bicycle infrastructure for Canada using open data sources
Bibliographic record
Abstract
High-quality and consistent cycling infrastructure data are needed to advance research into equity and safety and for planning active transportation. With recent growth in cycling and investments in cycling infrastructure, there are concerns that these investments have not been equitable across communities. There is no consistent and complete national dataset for cycling infrastructure in Canada. Our goal is to develop a national cycling infrastructure dataset by (1) classifying OpenStreetMap (OSM) using the Canadian Bikeway Comfort and Safety Classification System (Can-BICS) as consistent criteria and categorisation for comfort class and infrastructure type; (2) performing a site-specific accuracy assessment by comparing the classification with more than 2000 reference points from a stratified random sample in 15 cities; and (3) presenting summary results from the national dataset. Based on reference data collected in 15 test cities, the classification had an estimated accuracy of 76 ± 3% for presence or absence of infrastructure, 71 ± 4% for comfort class and 69 ± 4% (by length) for infrastructure type. High comfort infrastructure was slightly underestimated (since bike paths were sometimes confused with multi-use paths) and low comfort infrastructure was slightly overestimated. Nationally, we identified 22,992 km of cycling infrastructure meeting Can-BICS standards and 48,953 km of non-conforming infrastructure. Multi-use paths are the most common infrastructure type by length (16.6%), followed by painted bike lanes (11.0%), and then high comfort infrastructure (cycle tracks, local street bikeways and bike paths) (4.3%). There was a wider range in access to cycling infrastructure in small cities than in medium and large cities. To reduce repeated effort assembling data and increase reproducible active transportation research, we encourage contribution to OSM.
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 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.008 | 0.019 |
| Science and technology studies | 0.003 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.009 | 0.005 |
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".