Road-Data Acquisition
Road Selection and Context
Click a road, retrieve nearby public map data, preserve relevant OSM tags, and identify the selected road segment. Supports: selected-road context and export evidence.
Feature Deep Dive
Feature Deep Dive for the Live Model
Road Risk is not just a map demo. It is a live analytical workspace for selecting roads, inspecting geometry, changing assumptions, comparing distributions, analysing routes, and exporting evidence.
Capability Map
Analytical Functions Within the Live Application
Each feature supports the same public-data to risk-output chain. The app should be used as an exploratory model, not as a black-box verdict.
Geometry Extraction
Curvature and Radius Outputs
Estimate segment length, heading change, bearing, curvature, radius, and related values from selected road geometry. Supports: safe speed, curvature risk, and maths output.
Physics-Based Outputs
Safe-Speed and Stopping-Distance Checks
Translate speed, friction, curvature, reaction time, and stopping assumptions into readable physical outputs. Limitation: this is simplified physics, not full vehicle simulation.
Scenario Testing
Vehicle and Condition Assumptions
Change vehicle class, weather, lighting, behaviour, BAC, fatigue, distraction, speed, and road-context assumptions. Supports: sensitivity testing.
Infrastructure Context
Road Class, Tags, and Fallback Assumptions
Use highway class, speed tags, surface, smoothness, lanes, lighting, traffic proxies, and missing-data flags to keep context visible. Limitation: absent public tags do not prove absent infrastructure.
Statistical Comparison
Distribution-Based Comparison
Sample the visible network and compare selected-road output against percentile, density, and tail views. Supports: context beyond a raw model output.
Route and Isochrone Analysis
Spatial Analysis Outputs
Calculate routes, draw route lines, sample route risk, inspect hotspots, and explore isochrone-style reachability contexts. Supports: route mean and peak segment interpretation.
Export and Transparency
Reviewable Evidence Outputs
Export CSV, GeoJSON, JSON summaries, distribution data, and visual outputs so the result can be reviewed outside the app. Supports: reproducibility and audit.
Research and Education Use
Inspectable Modelling Workflow
Use the app as a teaching and research surface where assumptions, formulas, and limitations are visible rather than hidden.
Feature-to-Output Map
Feature Groups and Supported Outputs
| Feature group | Scientific purpose | Outputs supported | Caution |
|---|---|---|---|
| Road data and selection | Build the selected-road context from public geometry and tags. | Road info, selected segment, exports. | Parallel roads and missing tags require careful interpretation. |
| Geometry extraction | Convert road shape into measurable length, bearing, heading change, radius, and curvature. | Curvature, safe speed, maths panel. | Sparse or noisy road nodes can reduce confidence. |
| Scenario testing | Show how assumptions change model output direction and magnitude. | Annual output, daily view, safe speed, stopping distance. | Assumptions should be changed deliberately, not treated as observed facts. |
| Infrastructure context | Separate retrieved OSM tags from derived values and fallback assumptions. | Infrastructure rating, confidence notes, exported metadata. | Missing public tags should not be interpreted as proof that infrastructure is absent. |
| Statistics and graphs | Place one road inside a sampled comparison set. | Percentile, density, tail, distribution exports. | The comparison depends on the sampled area and available road data. |
| Route, isochrone, and export tools | Extend selected-segment reasoning to routes, reachable areas, hotspots, and review files. | Route mean, highest segment, isochrone overlays, CSV, GeoJSON, JSON, PNG outputs. | Route and isochrone outputs remain sampled model interpretations, not official assessments. |
Scientific Workflow
Recommended Analytical Workflow
01
Select a Road
Use the map to select a real road segment and verify that the highlighted geometry matches the intended road.
02
Inspect Assumptions
Review missing tags, fallback assumptions, vehicle profile, speed, weather, lighting, and traffic context.
03
Read Model Outputs
Compare Annualised Comparative Model Output, daily view, safe speed, stopping distance, infrastructure rating, and maths notes.
04
Compare Context
Open graph views or route analysis to understand whether the road is typical, elevated, or unusual in context.
05
Export Evidence
Use export formats to preserve geometry, tags, scenario assumptions, model values, and graph data for review.
Model Output Families
Model Values Beyond Visual Map Colour
- Risk
- Annualised Comparative Model Output, daily interpretive view, percentile position, and route-average model output where applicable.
- Physics
- Curvature, radius, safe speed, stopping distance, friction assumptions, and speed comparisons.
- Context
- OSM tags, road name/ref where available, surface, speed tags, traffic proxy fields, and fallback confidence notes.
- Evidence
- Maths panel, graph modal, map overlays, route hotspots, and export files all support review of the same selection.
Selected Road
Live
Risk card, maths, graph data, route overlay, and export state update from the active road context.
Scenario Controls
Vehicle, weather, visibility, behaviour, traffic, surface, and speed assumptions.
Analytical Factors
Geometry, Physics, Context, and Data Confidence
Road Geometry
OSM Ways, Nodes, Bearings, and Selected Segment
Coordinate geometry supports segment length, heading change, radius, and curvature. Sparse or noisy geometry lowers confidence.
Physics
Friction, Safe Speed, Slip Ratio, and Stopping Distance
Speed, radius, friction, reaction time, and vehicle assumptions create interpretable physical checks.
Scenario Assumptions
Weather, Visibility, Lighting, Behaviour, and Surface
Rain, fog, ice/snow, heat, fatigue, distraction, BAC, overspeed, and surface quality can change interpretation.
Infrastructure Context
Road Class, Speed Tags, Lanes, Lighting, and Traffic Proxy
OSM tags and fallback assumptions help describe context without pretending missing data is proof of absence.
Statistics
Raw Output, Percentiles, and Upper-Tail Markers
Sampled-network comparison, P95/P99 style thresholds, and distribution spread make outputs easier to interpret.
Routes and Exports
Route Mean, Highest Segment, Hotspots, and Evidence Files
Route overlays, sampled segments, CSV, GeoJSON, JSON, distribution data, and PNG outputs support review.
Research and Education
Capabilities Designed for Auditability
Maths
Equation Traceability
The maths page and modal explain how physical quantities connect to displayed values.
Graphs
Network Comparison
Distribution views show where one selected road sits relative to sampled roads.
Exports
Review Outside the Browser
Export formats make it possible to audit values, geometry, and scenario assumptions.
Limitations
Assumption Visibility
Fallbacks and missing public data are part of the interpretation, not hidden from the user.
Live Application
Use These Capabilities in the Live Application
The public site explains the system. The live app is where the map, model, graphs, routes, and exports run.