While human activities in cities alter local climates leading to increased risks of heat-related stress and mortality, urban residents are expected to surpass 70% of the total global population by 2050. Traditional heat exposure assessments rely on readily available estimates of residential populations. As such, they overlook the fact that different population groups have diverse activity schedules and mobility patterns, which – combined with the heterogeneity of urban environments – may significantly alter heat exposure and risk. To address this problem, we present a novel framework that leverages a city’s open data and state-of-the-art agent-based modelling to describe humans as active components of the urban systems. As a case study, we focus on interconnected small- to medium-size cities of Cantons Vaud and Geneva in Switzerland. Our framework consists of extensive data mining and processing to homogenize information on buildings’ forms and functions, land uses and covers, transport systems, population, and human activities (based on the Swiss TimeUse+ Survey, Winkler et al., 2024). We use the state-of-the-art model DAVE (Dynamic Anthropogenic actiVities and feedback to Emissions) (McGrory et al., 2024, Hertwig et al. 2025) which couples a behaviour model with a transport model, a building energy scheme, and a land-surface model to simulate people’s activity and location as well as the resulting anthropogenic heat emissions at 10-min resolution. This study demonstrates that using open data and a bottom-up modelling approach incorporating active human behaviour into urban system modelling is key to improving urban climate adaptation and mitigation strategies.