Distributed, Ambient and Pervasive Interactions Best Paper Award

Paper Abstract
In the current context of accelerating climate change, reducing urban CO₂ emissions has become a key strategic priority for cities worldwide. This study, developed under the ReGenWest project, proposes a replicable, data-driven framework to optimize the combination of renewable energy systems (RES) and nature-based solutions (NBS) for urban decarbonization in Thessaloniki, Greece. Eleven typologies of open spaces suitable for NBS and RES interventions were identified in the study area, confirming its suitability as a pilot for the project. The framework allows platform users to flexibly define the specific area of intervention based on stakeholder boundaries and local priorities as a single optimization block, reflecting the distinct ownership and decision-making context of each site. The methodology integrates NetLogo-based simulations with Python optimization tools, including an exhaustive search option, to evaluate feasible RES-NBS configurations at the scale of spatial fundamental decision units. The optimization problem was formulated and solved as a Multiple-Choice Knapsack Problem using Google OR-Tools. Results reveal cost-CO₂ trade-offs shaped by non-linear dynamics, arising from the incorporation of real-world constraints such as economies of scale, which produce non-linear Pareto curves that better reflect the complexity of real-world interventions. The framework provides a transparent decision-support tool for planning cost-efficient, spatially feasible interventions, replicable in other urban contexts. Future developments will incorporate additional non-linear constraints and AI implementation to enhance accuracy and support AI-driven optimization.

The full paper is available through SpringerLink, provided that you have proper access rights.