Illicit Nuclear Trafficking
Paul Whitney and Sean Kreyling
Executive Summary
Illicit nuclear trafficking networks are a national security challenge. These networks can directly lead to nuclear proliferation, as states or non-state groups attempt to identify and acquire nuclear weapons-related expertise, technologies, components, and materials. The ability to characterize these networks and improve our ability to anticipate the key nodes, and transit mechanisms associated with them, is essential to influencing, disrupting, destroying, or interdicting the function of the network and its processes.
Approach
Our contribution to addressing this challenge includes mathematics and computational collaborative methodology to forecast organization behavior using models, empirical data and experts' opinions. Our technical approach to integrate predictive information has three unique and significant aspects: rigorous model validation, formal modeling of experts' assessments and automatic support for model integration. Each is an essential step to improving the forecasting of organizational behaviors. We take an 'engineering' approach to validating the integrated forecasting model — executing a detailed quantitative comparison of dynamic behavior models to observations. We incorporate experts' assessments through a mathematical structure linking experts' input and the computational model. Our approach treats experts as 'sensors', incorporating their opinions through sensor model component. Finally, the diverse concepts and expertise to address these challenges drives developing automatic support for model integration. These three aspects involve significant traditional science and technology expertise, combined with detailed knowledge of extant social systems (e.g. banking, energy distribution, insurance...) and technical systems. Enabling and supporting integration of models, experts' opinions and data is a critical step forward to address the challenge of illicit nuclear trafficking.