[MOVING TOWARDS GREATER SAFETY]
Severity Assessment
QUANTIFY RISKS // PREVENT CRASHES // OPTIMIZE INFRASTRUCTURE
From Simulation to Human Consequence
Roadscor quantifies severity by mapping simulated conflicts to real-world medical outcomes. Developed with VUFO (TU Dresden) and the GIDAS database, our process anchors risk in observed physics rather than generic assumptions.
- Real-World Mapping: We extract kinematic descriptors descriptors – including relative speed, impact angle, vehicle mass, and road-user type (including age for VRUs) – to map simulated conflicts to statistically comparable real-world accidents.
- Probabilistic MAIS Outcomes: Using VUFO methodology, we determine the probability of Maximum Abbreviated Injury Scale (MAIS) outcomes, providing a distribution of injury levels from minor to fatal rather than a single deterministic guess.
- Non-Linear Weighting: A Fibonacci-based scheme ensures high-severity risks are appropriately emphasized, reflecting their true societal impact and preventing fatal risks from being "averaged out."
- Traceable Metrics: Each score remains fully traceable to the original route pairing and kinematic inputs, ensuring a transparent and defensible audit trail.
The Evidence-Based Bridge
By grounding our analysis in real crash outcomes, Roadscor bridges the gap between digital traffic models and human consequences. This approach aligns with the rigorous standards used in automotive development and automated-driving safety, enabling meaningful safety evaluation long before implementation on-site.