JSR

Cities are increasingly understood as complex adaptive systems composed of interdependent social, economic, environmental, and technological subsystems. Traditional smart city models—focused on isolated technologies—fail to address this complexity. This study introduces a conceptual and technical framework that redefines the smart city as a digital ecosystem, where autonomous agents represent urban actors—including citizens, infrastructures, services, and industrial entities. These agents operate within shared data spaces, common ontologies, and negotiation protocols, enabling distributed decision-making, adaptive coordination, and continuous cross-domain co-evolution. The framework was examined through three practical cases: (1) a multi-agent traffic management system in Taipei that improves flow efficiency and reduces emissions; (2) an adaptive multi-resource smart-grid model demonstrating self-organizing balancing of electricity, gas, and heat; and (3) a cross-domain integration scenario linking Smart City (SC) and Industry 4.0 ecosystems through semantic mediation and multi-agent negotiation. Results show that ecosystem-based urban management enables real-time optimization, interoperability, sustainability and resilience across sectors. The paper further introduces the Digital Ecosystem Sustainability Evaluation System (DESES)—a method for assessing sustainability as an emergent, adaptive property rather than a static collection of KPIs. The study shows that digital ecosystems shift sustainability and resilience from predefined policy targets to emergent outcomes of system design. This paradigm enables cities to self-organize, co-evolve with their environment, and respond adaptively to disruptions. The proposed framework offers a foundation for next-generation urban management systems that integrate people, technologies and governance within a unified, living digital ecosystem.

smart city; digital ecosystem; sustainability; resilience; multi-agent systems; ontologies; urban governance.