Modeling factors contributing to dockless escooter injury accidents in Austin, Texas

Abstract: This study aims to identify factors influencing e-scooter injury accidents in Austin due to concerns about rising ridership and insufficient accident data. Using 2018 dockless e-scooter injury data, we employed zero-inflated Poisson (ZIP) and zero-inflated negative binomial (ZINB) models. Results indicate the ZIP model better fits the data. Significant variables include the ratio of 18- to 34-year-old males to females, median annual household income, ratio of public to private transport users, land use entropy index, percentage of restaurants, and percentage of educational centers. Recommendations include infrastructure improvements in dense urban areas, a demerit point system for unsafe riders, and educational campaigns by e-scooter operators and law enforcement.