FISC-AF: Federated Interoperable Smart-City Architecture Framework – Simulation Data, Traceability Matrices and Evaluation Results for Loja, Ecuador
Description
This dataset supports the manuscript “Federated Architectures for Ethical and Interoperable Smart City Governance…” submitted to the Journal of Urban Technology. It contains the complete simulation data, traceability matrices, and evaluation results for the Federated Interoperable Smart-City Architecture Framework (FISC-AF). Contents: -High-fidelity hybrid discrete-event and agent-based simulation models (AnyLogic v.8.8) for mobility, energy, and public safety domains in Loja, Ecuador 10,000 Monte Carlo runs per scenario with 95% confidence intervals -Full KPI results (baseline vs. FISC-AF) including equity-weighted metrics for peripheral neighbourhoods - Complete traceability matrices mapping FISC-AF to TOGAF 10th Edition ADM, ISO/IEC/IEEE 42010:2022 viewpoints, and ISO/IEC 42001:2023 AI management system controls - Expert validation scores (16 TOGAF-certified architects + municipal leads) - Simulation assumptions, cross-validation statistics (MAPE, RMSE, R²), and sensitivity analysis Purpose: Enable full reproducibility of the Design Science Research evaluation and support adoption of FISC-AF by other mid-sized municipalities in the Global South. Software required: AnyLogic 8.8 (or higher), Microsoft Excel / LibreOffice Calc, and a standard PDF reader. The dataset is linked to the associated research article on federated enterprise architecture, ethical AI governance, and digital-divide mitigation in resource-constrained urban contexts. Keywords: federated enterprise architecture, smart city interoperability, ISO/IEC 42001 AI governance, TOGAF 10th Edition, digital divide mitigation, design science research, AnyLogic simulation, Monte Carlo methods, Loja Ecuador, ethical AI governance, urban equity, NIST IES-City PPIs
Files
Steps to reproduce
The dataset was generated through a Design Science Research (DSR) methodology aligned with the TOGAF 10th Edition Architecture Development Method (ADM). Four iterative build-evaluation cycles were conducted following the six-step DSR process proposed by Peffers et al. (2007). Problem identification and motivation Systematic literature review and analysis of Loja Municipality Ordinance No. 0068-2024 and Ecuador’s National Digital Transformation Policy 2025–2030 identified the need for a federated, standards-compliant smart-city architecture. Definition of solution objectives Objectives were defined to deliver semantic, syntactic, and organizational interoperability, embed ISO/IEC 42001:2023 AI controls, and mitigate the digital divide through dedicated viewpoints (Citizen Experience and Andean Sustainability). Design and development The FISC-AF artifact was constructed through four iterative cycles mapped to TOGAF ADM phases (Preliminary to Phase H). Each cycle produced layered architecture descriptions, 12 ISO/IEC/IEEE 42010:2022 viewpoints, McKinsey-inspired federated governance structures, and embedded ISO/IEC 42001 risk-treatment controls. Full traceability matrices were maintained throughout. Demonstration The framework was instantiated in three interdependent domains (mobility, energy, and public safety) using Loja’s operational LoRaWAN infrastructure and fully anonymised 2023–2025 datasets. Evaluation High-fidelity hybrid discrete-event and agent-based simulation models were developed in AnyLogic v.8.8. Each scenario was executed with 10,000 Monte Carlo runs. Statistical validation included paired t-tests, ANOVA (p < 0.01), false discovery rate correction, 70/30 temporal cross-validation, MAPE, RMSE, and R² metrics. Sensitivity analysis (±20% parameter variation) confirmed robustness. Sixteen TOGAF-certified architects and two municipal ICT leads performed structured expert validation (average score 4.8/5.0). All simulation models, traceability matrices, KPI calculation tables, and expert review instruments are included in this dataset. To reproduce the results: Open the AnyLogic models in AnyLogic 8.8 or higher. Load the provided parameter files and run the Monte Carlo experiments (10,000 replications recommended). Use the Excel files to replicate KPI calculations and equity-weighted analyses. Consult the traceability matrices to verify standards compliance. The complete statistical code, simulation assumptions, and cross-validation procedures are documented in the supplementary files.
Institutions
- Universidad Politécnica de MadridMadrid, Madrid