The events of 9/11 led NASA to introduce a security aspect to its Aviation Safety Program with a portfolio of a dozen advanced technologies designed to make flight more secure. A complete prioritization and management of the portfolio was required to ensure maximizing investment payoffs in the highest promising technologies and to perform a technical integration of the various NASA technologies with each other as well as within the overall National Airspace System (NAS).

To accomplish this, a new approach was required. A complete vulnerability assessment had to be performed of the NAS to formulate a baseline risk, then the NASA technologies could be applied to find the risk reduction provided. The enormous task of performing a risk-based analysis of the NAS was best managed by using a computational methodology called the Logic Evolved Decision (LED) analysis developed at the Los Alamos National Laboratory for modeling the behavior of complex systems with respect to decision support applications. The LED computational algorithm uses linked, formal logic models to represent the basic functions of a decision analysis tool. This systems analysis capability incorporates fuzzy logic, approximate reasoning, possibility, probability, multi-attribute scoring, and graph theory to construct these decision support models. LED is a flexible, self-contained, comprehensive, and traceable decision support software tool for risk-based prioritization and portfolio management across a broad spectrum of applications.

Using the LED tool, a comprehensive set of several million attack scenarios was developed using fault tree analysis to define terrorist threats to aircraft, airports, and the airspace. Then a much smaller representative spanning set of attacks were chosen and used to apply the technologies on to determine the effective risk reduction. To find the effective risk reduction, factors such as the technology readiness levels, technical development risks, implementation risks, cultural and certification issues, and cost/benefit had to be considered. The power of using the LED methodology is the use of approximate reasoning in constructing inference models to analyze each attack scenario. Through the use of expert elicitation for the attack scenarios and inference models, NASA technologists, experts in all aspects of aviation operations and security, and national security analysts could input their expertise to the threats, the security technologies, and the impact of the technologies. From the application of these inference models to the attack scenarios, a final risk could be found and compared to the baseline risks to determine an effective risk reduction. These risk reductions were analyzed in a variety of ways. Primarily, they were looked at as each technology was used in a "stand alone" operation, or coupled with one or more other technologies for increased risk reduction.

The objective of these assessments was to provide a decision support tool that can be used to prioritize NASA research in aviation security. This top-down analysis and modeling approach utilized the LED methodologies and techniques to rank order the proposed NASA security research projects by using risk reduction as the metric.

Glenn Author and Contact:
Kenneth L. Fisher, 216.433.5655, Kenneth.L.Fisher@nasa.gov

Programs/Projects:
Aviation Safety Program