SciFest road-safety research

Public model: 2.0SciFest 2026

Inspect how road geometry and assumptions change a comparative model output

Road Risk turns public road geometry and explicit scenario settings into an inspectable research result. It is built for comparison and method review, not for declaring a road safe or predicting a collision.

Open Road Risk App Start Judge Review

Current App

See a real road calculated section by section

This frozen R238 example uses OpenStreetMap centreline geometry and the same Model 3.2 calculation, local sections, Maths, Graphs and export paths as the App.

Road Risk App analysing Malin Road, R238 in County Donegal with Model 3.2. The map shows 42 hard-edged colour-coded sections, the expanded output card, quantitative legend, tool dock and OpenStreetMap attribution. — **Demonstration boundary:** Real OpenStreetMap centreline geometry, frozen for repeatable demonstration. It is not a surveyed lane alignment, not a collision record, not a safety rating and not an empirical validation case. Green means lower comparative output on the fixed display scale; it does not mean safe. Road geometry © OpenStreetMap contributors, ODbL. Basemap attribution remains visible in the capture.

What the project does

Four connected research tasks

Inspect a road

Select public road geometry and examine the current model output, local sections and source support.

Compare routes and areas

Sample routes, local areas and reachability analyses under one visible scenario.

Change assumptions

Hold a road constant while changing weather, surface, behaviour or vehicle settings.

Review the evidence

Open formulas, calculation traces, geometry diagnostics, sources, warnings and exports.

Status and boundaries

Current software and current evidence are reported separately

Current model status

Model 2.0.0 is the public default

Compare model versions

Model 2.0

Public default

Empirical decision: not assessed

Model 3.2

Controlled preview

Empirical decision: insufficient evidence

Model 3.3 geometry candidates

Diagnostic only

Empirical decision: insufficient evidence

Software verification and deployment status are reported separately from empirical evidence. No model output is an official safety rating or collision prediction.

Follow the national evidence study

Choose a review path

Start at the level that matches your visit

Public overview

Read the research question, scope, method and present limitations before opening the App.

Read the project story

SciFest judge review

Follow a timed five-step route through the question, method, evidence, demonstration and authorship.

Open the five-minute review