Assessing comprehensive impact of autonomous electric taxis via an integrated modeling framework

Autonomous electric taxis are a crucial technology for the sustainable urban transport transition, yet the multidimensional impacts of large-scale deployment and differentiated policy effects remain insufficiently assessed. This paper proposes a framework integrating BERTopic, Bass diffusion, parameterized binary Logit and system dynamics to analyze autonomous electric taxis’ penetration in the on-demand mobility market of Guangzhou, a representative Chinese megacity, and their accident-related economic, environmental and employment impacts. Results show that autonomous electric taxis will reach approximately 70% penetration by 2050, and save roughly 88.4 million yuan in socioeconomic costs from fewer traffic accidents. However, new travel demand will lift cumulative carbon emissions by around 760,000 tons and displace about 61% of traditional mobility jobs. Moreover, policy trade-offs emerge: price subsidies best promote penetration and promotional campaigns best lower accident costs, but both exacerbate carbon emissions and employment impacts; opening operating areas has limited market effects yet mitigates environmental impacts.