Shifting skies: A cross-country investigation of evolution of public perception toward urban air mobility through Twitter (X) discourse

Urban air mobility (UAM) is increasingly being recognised as a promising response to the challenges of rapid urban expansion and its negative externalities. While technological advancements in vertical take-off and landing (VTOL) aircraft have accelerated development in this space, the widespread adoption of UAM services hinges on societal acceptance driven by public perceptions. Understanding these perceptions, especially their variation across regions and over time, is critical for developing policies to maximise their adoption rate. This study leverages a large-scale and long-term Twitter dataset to discern the spatio-temporal evolution of public perceptions towards UAM. To this end, we employed a combination of machine learning (ML) and a large language model (LLM) for performing sentiment classification. Subsequently, sentiment polarities are integrated with time series analysis, indicating the prevalence of positive perception for most of the last decade, while detecting the effect of various real-world events. In terms of spatial K-means clustering results, it reveals four clusters of countries with distinct characteristics. For example, people in countries like the USA and Australia are observed to be highly opinionated towards UAM, while public discourse in Germany and India is more neutral. Finally, dynamic topic modelling coupled with an LLM-based representation uncovers underlying themes of public discourse. Topic model findings underline three major global themes: (1) industry innovation and testing, (2) unmanned aviation systems, and (3) mobility benefits. Furthermore, we identified in some cases that local themes driven by specific incidents have a more substantial effect in shaping the preferences than the generic global ones. The paper hence contributes to the literature by providing the first global-level dynamic spatio-temporal assessment of future UAM services. The insights are expected to offer valuable policy guidance for policymakers, regulators, and industry stakeholders aiming to improve the public acceptance of UAM technologies and consequently the uptake. • First dynamic spatiotemporal assessment of public perception related to Urban Air Mobility (UAM). • Utilises 12-year long (2010–2021) geotagged Twitter dataset across 6 countries. • Applies Large Language Models (LLMs) for robust sentiment analysis and dynamic topic modelling. • Major global themes include industry innovation, unmanned aviation, and mobility benefits. • Local incident-driven discourse may influence public perception more than generic global ones.