InterJournal Complex Systems, 1627
Status: Accepted
Manuscript Number: [1627]
Submission Date: 2006
Structure and Dynamics of Multiple Agents in Homeland Security Risk Management
Author(s): Gary Nelson

Subject(s): CX.44



ABSTRACT Homeland security risk-management invokes problems at the extremes of uncertainty and resource allocation. The use of probabilistic decision making and the market (including actuarial indemnification) are inadequate. But there is also an evident evolutionary approach to the problem, since every complex system we see is a successful competitor for limited resources under a large spectrum of contingencies. It is vital to discern the structural and dynamical principles of such evolutionary systems, extended from the strict biological process to human organizations and artifacts. The extension to risk management must focus on the logical, information-based, organization of multiple agents (decision makers) prospectively dealing with risk. There are many organizations (e.g., air traffic management) that embody continuous, hierarchically-scaled, risk management. Organizations intuitively adopt evolutionary structure, but there are many pathologies—both structural and dynamical—of organizational design. The opportunity exists to apply recent technologies of artificial-agent ecologies and open-communication systems in hybrid human/artificial risk-management systems. This application needs a formal understanding of the structure and dynamics of complex-adaptive systems (CAS) to avoid the human- organizational pathologies before we extend them to the hybrid systems. This paper: 1) Summarizes the structural and dynamic principles of CAS in risk-management terms; 2) Examines “resilience” as the descriptor of risk-managing CAS, and; 3) Applies hybrid human/artificial multi-agent systems (MAS) in a risk-managing, resilient, network. Keywords Complex Adaptive Systems, Multi-Agent Systems, Resilience, Risk Management, Homeland Security

Retrieve Manuscript
Submit referee report/comment

Public Comments: