Simulation models based on Artificial Intelligence for Sustainability and climate change actions - SAIS

Abstract

Climate actions and sustainability are key challenges we face today. In this project, the goal is to use agent-based models together with artificial intelligence (AI) techniques, a social bottom-up simulation paradigm, to represent and actuate over a set of real problems related to sustainability and climate change. These problems are diverse and linked to the knowledge and expertise of the international collaborators belonging to the project: from the adoption of eco-innovations in the maritime industry to over tourism problems and climate change agreements by governments and stakeholders. To cope with these problems, we propose to enrich agent-based models and evolutionary games with explainable AI and machine learning algorithms such as reinforcement learning and fuzzy linguistic rules. 

Thanks to these explainable and AI enriched agent-based models, we will be able to represent these social and environmental dilemmas and adoption of eco-innovations in a more accurate and valid way. The validity of these models will be also ensured by adding real data and by building the models with respect to well established theories. Additionally, we will use evolutionary optimization algorithms to provide with the best policies to solve the dilemmas and societal problems by connecting the optimization algorithms with the agent-based simulations. Therefore, SAIS project looks for novel AI-based simulation systems to tackle sustainability problems while providing with managerial policies and guidelines to stakeholders to alleviate them.