This comprehensive report underscores the interdisciplinary and data-driven nature of the PHOEBE project, advancing urban road safety modelling by integrating human factors, dynamic risk assessment, and traffic simulation within a unified framework, validated and demonstrated in diverse European urban contexts. Deliverable 3.2 closes Work Package 3 activities within the PHOEBE project, focusing on the development, refinement, and integration of the components in the PHOEBE framework. They are: Mode shift and induced demand Human behaviour Risk assessment Traffic simulation For each component, models were developed or enhanced to estimate the safety impact of changes better and capture human decision-making processes in urban mobility contexts, with a particular emphasis on mode shift, induced demand, and active travel behaviours. This deliverable document presents the finalised versions of the models developed through iterative technical enhancements and empirical validation across the three PHOEBE use cases: Athens (EL), Valencia (ES), and the West Midlands (UK). Each model was encoded into the Aimsun Next simulation environment and tested using scenario-specific interventions. The framework's integration involves structured internal data flows, with scripts and APIs developed to process microsimulation outputs (speed, flows, trajectories) into formats suitable for demand, behaviour, and risk models. Aggregation and transformation steps align data resolutions across components. The integration flow is organized into setup, simulation, and refinement stages, facilitating iterative model runs and indicator calculations. User guides and performance indicator definitions support framework operationalization. PHOEBE provides a comprehensive set of Key Performance Indicators (KPIs) across four categories: health and safety, mode shift, environment and operations. Deliverable 3.2 successfully finalizes and demonstrates the PHOEBE framework's components across pilot cities, showcasing advancements in predictive urban safety assessment with a focus on VRUs. The integration of mode shift, human behaviour, risk assessment, and traffic simulation models provides a robust foundation for evaluating safety impacts of urban interventions. Future work will focus on automation, scalability, user guidance, and stakeholder engagement to facilitate broader adoption across European cities