The goal was to create a dynamic road risk index, quantifying the influence of other road users’ behaviour on a given road segment.
Our Approach:
- Data Acquisition – Analysed over 1 billion telematics records from 53,000 vehicles across 39 industry sectors, collected by GPS-GNSS enabled EROAD devices.
- Fatigue Index – Estimated driver fatigue based on median trip time from last rest location to each road segment, separated by light and heavy vehicles, normalised, and clustered into five fatigue-risk categories.
- Frustration Index – Measured curvature ratios for road segments, and compared speeding rates for corner-to-straight versus straight-to-straight events. Identified greater over-speeding tendencies on straighter roads, especially for heavy vehicles on functional class 1 routes.
- Familiarity Index – Calculated proximity of harsh braking and speeding events to trip endpoints, normalised by trip length, showing higher harsh braking risk closer to destinations.
- Composite Risk Index – Combined the three indexes (fatigue, frustration, familiarity) into a single normalised value, clustered into five risk levels.
- Routing Demonstration – Compared routes between Taupō and Rotorua, showing that a shorter route could present 45.8% higher cumulative risk despite being 23.9% shorter in distance.
Key Findings:
- Risk Variability by Road Class – Functional class 5 roads generally showed lower risk; classes 1–3 displayed multi-modal risk distributions influenced by speed limits and driving behaviour.
- Behavioural Correlation – Heavy vehicles exhibited significantly higher over-speeding after long straight segments, while light vehicles’ speeding patterns varied by local road geometry.
- Close-to-Destination Risk – Nearly half of harsh braking events occurred within 8.6 km of trip endpoints, aligning with prior crash proximity research but with higher spatial precision.
Impact:
- Evidence-Based Planning – Provided government agencies with a population-based risk measure for infrastructure investment and operational policy.
- Vision Zero Alignment – Offered a framework for proactive safety interventions based on real-world aggregated driving behaviour.
- International Recognition – Findings were published and presented at international transportation research conferences, positioning EROAD as a leader in dynamic safety analytics.
