Agent-Based Distributed Component Services in Spatial Modeling

Agent technologies have been increasingly applied to spatial simulation and modeling in silico. Where multi-agent systems have been used for spatial simulation, agents have tended to be deployed as spatial objects in order to study emergent patterns from micro-level behaviors. Many of these applications only deploy a weak notion of agency. More recently, the concept has emerged in the spatial domain that agents can be deployed as services to assist in complex modeling tasks. Agent-based distributed component services bring a stronger notion of agency to spatial modeling and are particularly suited to achieving interoperability in heterogeneous computational environments. Two case studies are presented. In the first, agent-based services are deployed over a network for spatial data quality analysis. In the second, a variogram agent component is used to demonstrate how a collaborating multi-agent system can provide intelligent, autonomous services to carry out complex operations.