Year
2016
Abstract
When developing and negotiating systems to monitor and verify a nation’s weapon stockpile, policy makers must be able to evaluate the effectiveness of each alternative implementation against system objectives. Such objectives include: confidence that a treaty accountable item (TAI) initialized by the monitoring system is as declared, confidence that there is no undetected diversion from the monitoring system, and confidence that a TAI is dismantled as declared. To enable evaluation of these objectives, we have developed and implemented a discrete-event simulation (DES) model, which seeks to capture all processes and decision points associated with the progressions of virtual TAIs, with notional characteristics, through the monitoring system from initialization through dismantlement. The DES takes, as input, a complex scenario and produces, as output, data in the form of truth reports (e.g., actual TAI movements), observation reports (e.g., technology- and inspector-observed TAI movements), and declaration reports (e.g., host-declared TAI movements). The design of our model provides a platform for rapid “what-if” analysis by encoding many features of the underlying system as tunable parameters. In other words, we have built a decision-support tool that provides necessary data for both the evaluation of preconceived scenarios and the user- enabled exploration of “better” ones. The actual evaluation methodology is part of a post- processing phase, which is beyond the scope of this paper. In this paper, we demonstrate how the DES model can be used to build a simulated, notional warhead monitoring system, from the overall concepts of operations (CONOPS) to the accompanying technologies and their inherent error rates. We focus on host declarations and the inspections they precede, showing how unintentional mishaps over time may play a role in expected system performance.