Deriving functional safety (ISO 26262) S-parameters for vulnerable road users from national crash data

Area of Science:

• Road safety engineering
• Automotive functional safety
• Traffic accident analysis

Background:

• Advanced driver assistance systems (ADAS) and automated vehicles (AV) operate alongside vulnerable road users (VRU) in shared road spaces.
• Assessing functional safety (ISO 26262 S-parameters) is critical for minimizing crash frequency and injury severity involving VRUs.

Purpose of the Study:

• To develop a data-driven method for estimating injury severity (S-parameters) for four VRU types (pedestrians, bicyclists, scooterists, motorcyclists).
• To enhance population-based accident data (DESTATIS) with in-depth study data (GIDAS) for detailed injury severity mapping to the S-scale.

Main Methods:

• Transformed injury severity from DESTATIS (4-level scale) to ISS breakpoints (ISS ≥{4, 9, 16}) using GIDAS data, enabling translation to the S-scale.
• Analyzed crashes by VRU type, vehicle type (BTV, LTV), injury mechanism, and traffic domain (speed limit clusters).
• Calculated S-parameters (90th-percentile injury severity) with a one-sided 95% confidence level for each VRU, mechanism, and domain.

Main Results:

• Established a method to translate police-reported injury severities to the S-scale, facilitating the use of population data for 'severe injury' (MAIS 3+) proportion estimation.
• Derived S-parameters specific to VRU type, vehicle interaction, and traffic environment.
• Demonstrated applications in evaluating AVs, ADAS, and protective gear for VRUs.

Conclusions:

• The developed data-driven approach provides essential S-parameters for functional safety assessments in ADAS and AV development.
• Accurate injury severity estimation is vital for improving the safety of vulnerable road users in mixed traffic scenarios.
• This methodology enables better utilization of large-scale accident databases for safety-critical automotive applications.