Modelling crash propensity of carshare members

Area of Science:

• Urban planning and sustainable transportation
• Traffic safety and behavioral analysis
• Insurance risk assessment

Background:

• Carsharing systems offer sustainable urban mobility solutions.
• Reducing crash and insurance costs is crucial for carshare viability.
• Understanding user behavior in crashes is key to improving safety.

Purpose of the Study:

• Characterize carshare users involved in crashes.
• Identify factors influencing at-fault and not-at-fault driving.
• Explore quasi-induced exposure in carsharing.

Main Methods:

• Utilized data from GoGet carshare users in Sydney, Australia.
• Analyzed user demographics, driving habits, and crash involvement.
• Compared distributions of at-fault, not-at-fault, and general carshare members.

Main Results:

• Infrequent users, multi-car owners, and experienced drivers had fewer crashes.
• Weekend non-users, low-mileage drivers, and Australian license holders were more likely to be at fault.
• Significant differences observed between not-at-fault drivers and general members in driving patterns and car usage.

Conclusions:

• Incentive structures based on driving experience and insurance excess can enhance carshare safety.
• Targeted interventions can reduce crash-related costs for carsharing systems.
• Behavioral factors significantly influence crash risk and fault determination in carsharing.