Agent-Based Simulation and Assessment of NAS Security and Safety

The key innovation proposed here is the use of agent-based modeling and simulation to evaluate the safety of the National Airspace under crisis operations and develop tools for real-time planning, scheduling, and resource allocation decision aids for crisis management. We view the problem as one of simulating and controlling the emergent behavior of autonomous agents (aircraft and air traffic service providers in this case) in crisis situations. We propose to use NASA's agent-based Airspace Concept Evaluation System as the modeling framework into which we will integrate our models. We propose to evaluate the impacts of these malicious agents on the safety of NAS by using simulation to assess short term and long term NAS-wide safety impacts in terms of loss of separation, near misses, collisions, re-routes, controller workload, and economic impacts. The agent-based system will provide a real-time planning, scheduling, and resource allocation decision aid to be used for crisis management, by providing the user capabilities to develop and execute playbooks that represent various policies. Finally, we propose to develop safety metrics that will provide command center traffic management coordinators indicators to predict off-nominal activities in the airspace.