ASCENT: A Network-Wide Simulation of Air Traffic Flow Management Incorporating Airline Decisions William D. Hall and Erica H. Peterson * Abstract This paper describes the Advanced Systems Concept Evaluator for Novel Technologies (ASCENT) simula- tion test-bed. In order to capture effects important to collaborative air traffic management tool design, AS- CENT employs optimization to model the response of the airlines to air traffic management actions. The simulation is able to assess not only traditional mea- sures of system performance such as airport through- put and passenger-minutes of delay, it also measures the value to the airlines of their realized plans. AS- CENT allows one to compare the airlines’ optimal solutions under a proposed ATM tool to those un- der the present-day system, answering the question, “How much better could the airlines operate under the new technology if it were implemented?” AS- CENT has been applied both to the benefits analysis of a proposed Air Traffic Flow Management method and to an analysis of the Collaborative Arrival Plan- ner (CAP) under development by NASA. 1 Introduction Over the past decades, capacity of the air transporta- tion system has increasingly often fallen short of air traffic demand, resulting in delay. It has been esti- mated that such delay within the domestic air trans- portation system increased the operating costs of the U.S. airlines by $2.5 billion in 1995 [3], and according to the Air Transport Association [1], by $2.9 billion in 1998 and by $3.7 billion in 2000. These costs have * whall@draper.com and epeterson@draper.com; Charles Stark Draper Laboratory; 555 Technology Square, MS 3F; Cambridge, MA 02139; USA grown nearly monotonically over time as growth in demand out-paced growth in capacity. Planned air- port construction and expansion projects will not in- crease capacity commensurate with the latest FAA and NASA projections of demand growth. In an attempt to reduce the adverse impact of the growing capacity-demand imbalance, researchers and practitioners have been studying ways to use exist- ing capacity to better advantage. A number of these efforts have focused on methods that allow the users of the airspace system more control over operations, with the philosophy that the users hold key infor- mation necessary to make efficient control decisions. Future design and development of such collaborative air traffic management is envisioned to encompass in- creasingly complex system issues, such as en-route resource allocation and multi-center traffic manage- ment. Effective benefits analysis and design of such complex collaborative mechanisms requires tools to predict their performance before investment is un- dertaken to implement them. Because of the collabo- rative nature of the system, these tools must account for the range of user decisions that could arise from the variety of user objectives and constraints present in the system. 2 ASCENT Overview ASCENT models the behavior of the air transporta- tion system, including airline response to Air Traffic Flow Management (ATFM) actions. The simulation contains models of Air Traffic Management, models of the airlines and models of aircraft movement, turn- around and other processes, as shown in Figure 1. During a simulation experiment, the allocation of 1