JavaScript is Disabled
Your browser's JavaScript functionality is disabled. It has to be enabled to use this function of ConfTool.
Here you can find information on how to enable JavaScript
If you have any problems, please contact the organisers at info@hepa2026.si.

Conference Agenda

Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).

Daily Overview

Session

WR2_2_1: Abstract presentations: Active Living in Natural and Urban Environments
Location: White room 2
Session Chair: Berry van Holland

Presentation 4

10:25am - 10:35am

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

Kai Michael Gensitz¹, Shawan Mohammed², Daniela E. Ströckl³, Marc Augustin⁴, Claudio R. Nigg¹, Ciara McCormack⁵

1: Department of Health Science, Institute for Sport Science, University of Bern, Switzerland; 2: Chair for Distributed Signal Processing, RWTH Aachen University, Germany; 3: Faculty Engineering & IT - IARA, Carinthia University of Applied Sciences, Austria; 4: Social Medicine, Protestant University of Applied Sciences, Bochum, Germany; 5: Department of Sport Science and Nutrition, National University of Ireland, Maynooth, Ireland

Purpose: While artificial intelligence (AI) integrates rapidly into mobile health (mHealth) for physical activity (PA), its deployment in urban areas remains under-scrutinised. This scoping review maps AI architectures in mHealth for PA, operationalised theories, system inclusivity, ethical standards, and environmental sustainability.

Methods: Following JBI and PRISMA-ScR guidelines, seven databases were searched for urban, smartphone-based AI systems for adult PA interventions since 2011 (N=2,222). We deploy an AI-assisted pipeline on local servers. For rigorous validation at title/abstract screening stage, a human-screened 10% calibration sample (n=223; inter-rater reliability PABAK=0.53) underwent expert adjudication, driving 20 iterations of prompt optimization (RISEN framework). A fully unseen 5% test sample (n=110; PABAK=0.75) validated inference performance, achieving PABAK of 0.86, 100.0% (95% CI 78.20-100%) sensitivity, and 91.6% (95% CI 84.10-96.30%) specificity, saving 98% of human screening time. This calibrate-test architecture will be repeated for full-text screening and data extraction stages – final analyses in April 2026. Accuracy, time savings, and resource usage are logged for each stage; data screening specific metric is concordance (95% CI) – final analyses in April 2026. Targeted citation searching ensures saturation.

Results: Human written findings will detail: (1) AI architectures for PA assessment and promotion (e.g., passive sensing, hyper-personalised feedback); (2) encoding of psychosocial determinants within algorithms; and (3) inclusivity considerations, ethical compliance, and sustainability levers (e.g., on-device computing).

Conclusions: Methodologically, this review demonstrates a resource-efficient AI screening pipeline. Practically, it provides a foundational evidence map to align future urban mHealth PA interventions with Planetary Health and ethical standards.