Sharper curvature should generally increase comparative output under the same scenario.
Validation and Reliability
What Has Been Checked, and What Has Not
The project includes repeated scenario tests and logical consistency checks. It does not yet claim professional validation or full calibration against national collision-history datasets.
Validation Status Matrix
Current Reliability Position
| Validation Area | Current Status | What It Means |
|---|---|---|
| Internal consistency | Done / ongoing | Core directions such as curvature, radius, speed, friction, and stopping distance behave as expected. |
| Repeated scenario testing | Booklet-supported | Controlled assumption changes were used to check explainable model response. |
| Export evidence | Supported by app | Case JSON, CSV, GeoJSON, graph data, and local saved cases can document model outputs. |
| Collision-history calibration | Not completed | The model is not calibrated against full official collision datasets. |
| Field inspection | Not completed | True surface, visibility, drainage, signage, and defects require real-world inspection. |
| Professional engineering review | Future work | Operational use would require expert review and formal validation. |
Internal Consistency
Sanity Checks Expected from the Physics
Smaller radius should reduce the friction-limited safe-speed estimate.
Higher speed should increase stopping distance non-linearly because the braking component scales with velocity squared.
Lower effective friction should reduce safe speed and increase braking distance.
Adverse weather, visibility, and behaviour assumptions should increase output for the same selected road.
Percentiles should be interpreted only within the sampled comparison set.
Stronger Validation Plan
What Stronger Validation Would Look Like
Collision Data
Compare Against Official Records
Test whether higher model outputs align with trusted collision-history datasets once ethical and data-quality conditions are met.
Known Sites
Check Known High-Risk Bends
Compare selected outputs with locations already identified by professional or public-sector processes.
Engineering Review
Compare Against Audit Expectations
Ask road-safety or transport professionals whether model mechanisms match field-audit reasoning.
Field Data
Verify Surface and Visibility
Inspect whether OSM tags, surface assumptions, sightlines, lighting, and barriers match real road conditions.
Sensitivity
Test Assumption Ranges
Run controlled sensitivity analysis across friction, speed, reaction time, visibility, vehicle, and traffic-proxy settings.
Repeated Testing
What the Booklet Reports
Repeated Scenario Testing
The same roads were tested under multiple controlled assumptions to check that outputs changed in explainable directions rather than randomly.
Multiple Road Forms
The booklet describes testing across straighter sections, more curved roads, and different surrounding contexts to compare model behaviour.
Graph and Percentile Context
The model uses sampled-network comparison so a selected-road value can be positioned relative to other roads under the same assumptions.
Export and Evidence Trail
The app supports exports intended to make geometry, scenario values, tags, and outputs reviewable outside the browser.
Submitted Case Boundary
Exported case evidence is not empirical crash validation. It is a model-output record that requires review before public display.
Not Yet Completed
Validation Work Required Before Operational Use
Professional validation boundary
The model has not yet been fully calibrated against national collision-history datasets. Outputs should be treated as comparative indicators. Further empirical calibration and professional review would be required before operational or policy use.
Collision History
Calibration Against Observed Records
Future work should compare outputs against official collision datasets where appropriate and ethically available.
Field Data
Road-Condition Verification
Surface wear, drainage, signage, visibility obstructions, and local defects require direct inspection or higher-quality datasets.
Expert Review
Transport and Engineering Assessment
Assumptions should be reviewed with road-safety, vehicle-dynamics, and infrastructure specialists.
Technical Reliability
API and Data Robustness
Open public APIs, missing OSM tags, and sparse geometry introduce uncertainty that must remain visible.