The air traffic controller and Maxwell's Demon

Abstract

The present world wide system of Air Traffic Control universally depends on an information exchange involving ground radar observations and cooperative application of rules and procedures by pilots in the air and controllers on the ground. The chief function of the controller is to keep traffic moving while "assuring" no conflicts or mid air collisions. In areas of dense traffic the controller "workload" can get very high and one becomes concerned about the system(controller is part of the system) becoming overloaded, with consequent decrease in system safety. The concept of controller or system workload is intuitively understood but how does one measure workload for the purpose of classifying system safety and achieving proper system design? One needs to define a model of the situation. Air traffic has previously been described by analogy with the random motion of gas molecules (Alexander and Graham-Orr) but these descriptions have left out from the beginning (by the random assumption) the essential feature of control. Associated with using the random gas model one assumes that controller workload is proportional to the number of conflicts the controller must resolve. Though predictions made from the random gas model check reasonably with some results obtained from computer simulations, doubts arise when one questions how a random model can describe the real world controlled air traffic situation with its obvious lack of randomness (Jones and Lutze) and one asks, more precisely, for the limitations of the random gas model. The purpose of this paper is to provide an alternate mathematical model for air traffic control; one in which the element of control is built in from the beginning thus overcoming a basic objection to the random gas model. The model is based on the recognition that whatever the controller "does" to achieve and maintain control, his work effort is perceived by an observer as a decrease in the entropy of the traffic, e.g., the traffic becomes more orderly. The controller functions in direct analogy with Maxwell's Demon. Faced with a disordered velocity and position distribution of aircraft in a control zone, the air traffic controller introduces order by supplying information so as to achieve a decrease in entropy. Controller workload is measured by the information (negentropy) he must supply.<>