A multiagent approach to managing air traffic flow

Abstract

Abstract. Intelligent air traffic flow management is one of the fundamental challenges facing the Federal Aviation Administration (FAA) today. FAA estimates put weather, routing decisions and airport condition induced delays at 1,682,700 hours in 2007 [18], resulting in a staggering economic loss of over $41 Billion [42]. New solutions to the flow management are needed to accommodate the threefold increase in air traffic anticipated over the next two decades. Indeed, this is a complex problem where the interactions of changing conditions (e.g., weather), conflicting priorities (e.g., different airlines), limited resources (e.g., air traffic controllers) and heavy volume (e.g., over 40,000 flights over the US airspace) demand an adaptive and robust solution.In this paper we explore a multiagent algorithm where agents use reinforcement learning to reduce congestion through local actions. Each agent is associated with a fix (a specific location in 2D space) and has one of three actions: setting separation between airplanes, ordering ground delays or performing reroutes. We simulate air traffic using FACET which is an air traffic flow simulator developed at NASA and used extensively by the FAA and industry. Our FACET simulations on both artificial and real historical data from the Chicago and New York airspaces show that agents receiving personalized rewards reduce congestion by up to 80% over agents receiving a global reward and by up to 90% over a current industry approach (Monte Carlo estimation).

Agent-Based Modeling and Simulation for the Design of the Future European Air Traffic Management System: The Experience of CASSIOPEIA

“…In general, they simulate ATFM on the day of the flight [1][5] [17]. For example, IMPACT [5] simulates the airlines' decisions when meteorological conditions affect flight scheduling, and those conditions result in a constraint on the capacity of some airports.…”

An Agent-Based Approach for the Design of the Future European Air Traffic Management System

“…In general, agent-based approaches have been successfully applied to model ATM systems [1] [2] [5]. However, the design of future ATM systems, as it is defined by the goals of the SESAR programme in Europe, presents new challenges in agentbased modeling and simulation such as: (1) modeling new decision levels in ATM systems (such as strategic decisions related to flow and capacity management with longer temporal horizons), (2) designing new representation methods to simulate at a larger scale (e.g., multinational geographic areas in Europe) taking into account limitations concerning existing data, and (3) creating new practical tools (more easily available to the research community) to support the development of new ATM studies.…”

Hierarchical simulation for complex domains

“…Multiagent coordination is an important aspect of many domains, such as air traffic control [2], Robocup soccer [1] and power plant operations [5]. A learning or evolutionary algorithm will often convert a once computationally intractable search problem into a feasible guided search.…”