Date of Award

Winter 2012

Document Type


Degree Name

Doctor of Philosophy (PhD)


Computational Modeling & Simulation Engineering


Modeling and Simulation

Committee Director

John Sokolowski

Committee Member

Frederic D. McKenzie

Committee Member

Andreas Tolk

Committee Member

Bruce Copeland


Used to describe some interesting and usually unanticipated pattern or behavior, the term emergence is often associated with time-evolutionary systems comprised of relatively large numbers of interacting yet simple entities. A significant amount of previous research has recognized the emergence phenomena in many real-world applications such as collaborative robotics, supply chain analysis, social science, economics and ecology. As improvements in computational technologies combined with new modeling paradigms allow the simulation of ever more dynamic and complex systems, the generation of data from simulations of these systems can provide data to explore the phenomena of emergence.

To explore some of the modeling implications of systems where emergent phenomena tend to dominate, this research examines three simulations based on familiar natural systems where each is readily recognized as exhibiting emergent phenomena. To facilitate this exploration, a taxonomy of Emergent Behavior Systems (EBS) is developed and a modeling formalism consisting of an EBS lexicon and a formal specification for models of EBS is synthesized from the long history of theories and observations concerning emergence. This modeling formalism is applied to each of the systems and then each is simulated using an agent-based modeling framework.

To develop quantifiable measures, associations are asserted: 1) between agent-based models of EBS and graph-theoretical methods, 2) with respect to the formation of relationships between entities comprising a system and 3) concerning the change in uncertainty of organization as the system evolves.

These associations form the basis for three measurements related to the information flow, entity complexity, and spatial entropy of the simulated systems. These measurements are used to: 1) detect the existence of emergence and 2) differentiate amongst the three systems.

The results suggest that the taxonomy and formal specification developed provide a workable, simulation-centric definition of emergent behavior systems consistent with both historical concepts concerning the emergence phenomena and modern ideas in complexity science. Furthermore, the results support a structured approach to modeling these systems using agent-based methods and offers quantitative measures useful for characterizing the emergence phenomena in the simulations.