Exploring Smartphone-Based Artificial Intelligence Systems for Physical Activity in Urban Areas: A Scoping Review with a Focus on Psychosocial Theory, Inclusivity, Ethics, and Sustainability

Over half of the global population resides in urban areas and a vast majority of adults worldwide own smartphones, which increasingly become pervasive sensing and intervention platforms, positioning human–computer interaction at the center of everyday behavior change (Harari & Gosling, 2023; Pawar, 2025; Ritchie et al., 2025). Building on this opportunity (An et al., 2023; Gao et al., 2024; Wang et al., 2023), we propose a scoping review that maps how digital health via smartphone (i.e., mobile health) uses artificial intelligence (AI) systems to assess and promote physical activity (PA) in urban areas (see Figure 1). The review will pay explicit attention to psychosocial theory integration, inclusivity and ethical considerations as well as sustainability aspects of AI systems. We will synthesize literature that spans mobile sensing, machine learning assessment, and AI-enabled interventions, while emphasizing design considerations for responsible, human-centered real-world deployment. The aims are to (1) chart the landscape of AI-enabled systems on smartphones as mobile health assessment (e.g., passive sensing, computer vision, natural language processing) and intervention (e.g., just-in-time adaptive interventions, on-device/in-cloud causal reasoning); (2) identify how psychosocial theory (e.g., Integrated Behavior-Change Model, Social-Cognitive Theory, Behavior Change Techniques) is operationalized in algorithms, interfaces, and user journeys; (3) reflect on surface equity, ethical considerations, accessibility practices (e.g., language, ability, context constraints), and existing gaps; and (4) articulate sustainability levers (e.g., energy-aware modelling, on-device vs. in-cloud computing). Following the PRISMA-ScR framework, we will implement a transparent, reproducible protocol with the following eligibility choices (Peters et al., 2022; Tricco et al., 2018): (a) Population: adults (18+), Concept: smartphone-based AI systems for PA - assessment and intervention - plus system development processes - psychosocial theory, inclusivity, ethics, and sustainability -, Context: urban areas globally; (b) Source databases spanning health and technology research (e.g., PubMed, PsycINFO, SportDiscus, Web of Science, Scopus, IEEE Xplore, ACM Digital Library) plus targeted citation searching on seed references (Hirt et al., 2024); (c) Timeframe includes studies from the first AI deployed on smartphones onward (i.e., Siri in 2011; Bosch, 2018) in English and German; (d) Data charting extracts methodological variables (e.g., design, sample, context) and outcome domains aligned with our conceptual framework: Technology (i.e., AI-type; on-device vs. cloud; sensing modalities), PA outcomes (e.g., exercise, daily activity, sedentary behavior disruption, active transport), psychosocial theory factors (e.g., intention, habit), inclusivity and ethics (e.g., accessibility, subgroup tailoring), sustainability (e.g., energy/climate considerations, nature-linked activity design), and system idea/development process (e.g., assessment or intervention, research vs. business, co-creation). Analysis will be descriptive (e.g., frequencies, evidence maps) with a conceptual synthesis of AI system designs and disseminations as mobile health via smartphones in urban areas. This review thus covers mobile health in applied contexts (i.e., PA promotion and behavior maintenance through AI via smartphones as everyday technology) and provides a rigorous map of current mobile health research, populations which are targeted, and conditions under which mobile health is implemented. A forward-looking agenda for informed, inclusive, ethical, and sustainable human–computer interaction in mobile health research and practice for the AI era concludes the review. References An, R., Shen, J., Wang, J., & Yang, Y. (2023). A scoping review of methodologies for applying artificial intelligence to physical activity interventions. Journal of Sport and Health Science. https://doi…