ad-dataset-analysis

To calculate crash rates for autonomous driving systems, this Skill analyzes crash databases like FARS, CRSS, GIDAS, and SHRP2. You input crash and disengagement data, match ODD conditions, and execute the 'calculate_crash_rates' command to perform exposure-based risk analysis.

For naturalistic driving studies, you can compute surrogate safety metrics including Time-to-Collision (TTC), Post-Encroachment Time (PET), and Deceleration Rate to Avoid Crash (DRAC). These metrics help evaluate traffic safety and potential improvements for autonomous vehicles.

Statistical significance testing for autonomous driving safety claims is performed using Poisson tests and required sample size calculations. This Skill utilizes Python libraries like scipy and statsmodels to validate safety demonstrations and compare relative risk between systems.

Yes, you can use pandas and geopandas to analyze autonomous driving crash data. The Skill requires Python libraries including pandas, numpy, scipy, statsmodels, and geopandas to process crash datasets and generate spatial safety metrics for risk assessment.

The best way to compare autonomous driving safety with human driving risk is by inputting crash and disengagement data, matching ODD conditions, and calculating relative risk. This Skill provides comparative safety analysis using crash rates and statistical significance testing.