SPADE as a MAS Framework for the Development of Digital Twins
摘要
This paper presents SPADE (Smart Python Agent Development Environment) as a framework for developing Digital Twins using Multi-Agent Systems (MAS). SPADE supports scalable and distributed architectures through its XMPP-based communication system, asynchronous programming model, and agent-based design. The SPADE Artifacts extension, based on the A&A (Agents and Artifacts) meta-model, provides a structured interface between agents and physical systems. To validate its effectiveness, we present a case study on urban traffic management in Valencia, Spain, where 386 street segments are modeled as agents that track real-time traffic, predict congestion, and coordinate redistribution. The system integrates diverse data sources and visualizes network conditions via an interactive dashboard, demonstrating SPADE’s potential for real-world Digital Twin applications.