CAREER: Evaluating Cooperative Intelligence in Connected Communities

This project explores a vision where groups of buildings autonomously cooperate to share energy resources such as solar panels, batteries, or generators. Through multi-agent simulations and real-world experiments, it aims to reveal how cooperation can emerge and persist within self-organizing energy systems, and how it can lead to collective benefits such as energy savings and climate adaptation. By harnessing collective efforts, this work aims to catalyze a shift towards collective infrastructure systems. Beyond its targeted scientific contributions, the project includes an interactive educational program featuring games and simulations fostering public understanding of cooperation as a vital aspect for energy equity and sustainability. The goal is for findings to empower architects, planners, and policymakers with knowledge and tools to design and manage connected communities where buildings share, rather than compete for, energy resources. The project aspires to inspire a new way of thinking about how collective collaboration can create better energy systems.This project will investigate how cooperation can emerge and be sustained among groups of buildings that strategically pool distributed energy resources (DERs) to enhance shared resilience and sustainability. It will establish an interdisciplinary framework to investigate cooperation within multi-agent energy systems. The technical approach synergistically combines reinforcement learning-based control policies, multi-agent simulations, and empirical validation. It seeks to formally model the complex reciprocal interactions underlying cooperative energy behaviors within connected communities. The overarching goal is to elucidate quantitative design principles and policy guidelines that embed cooperative intelligence within engineered, self-organizing architectures to enhance collective resilience and equitable resource allocation. The first thrust of the project involves developing an adaptable simulation testbed to evaluate cooperation trajectories under various energy management schemes, community configurations, and decision protocols. This flexible virtual testbed will enable systematic evaluation of factors that influence emergent cooperation. The second thrust entails an empirical analysis to diagnose how real-world socio-technical elements—from human preferences to technology diffusion asymmetries—shape cooperative outcomes. Grounding insights in data will uncover actionable pathways for implementation. Finally, interactive web platforms and games based on constructivist learning will cultivate public understanding of decentralized systems and cooperative intelligence for equitable and resilient energy access. By integrating insights across technical and social dimensions, this project will inform the design of efficient, effective connected communities and foster sustainable, equitable energy systems. By elucidating strategies for localized sharing and energy management, the project will showcase models of distributed ownership and governance.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.