An Experimental Investigation of Dynamic Allocation of Tasks between Air Traffic Controller and A.I. Systems

Abstract

Abstract The air traffic increase and the air traffic controller workload heaviness lead to provide an assistance to the air traffic controller. The purpose of our research is to propose and validate a new organization of the air traffic control, which allows air traffic controllers to stay active in the control and supervisory loop of the process, in order to maintain the present traffic safety level and to improve the global system performances. Our research consists in decomposing the problem according to the two levels of the air traffic control organization : a tactical level which aims at regulating the traffic and a strategic level which aims at filtering this traffic. The first step of our research is directed towards an horizontal cooperation that consists in a dynamic allocation of the tactical level control tasks between human air traffic controllers and an assistance tool. This paper presents the dynamic tasks allocation principles, and then describes the experimental platform for tasks allocation in air traffic control. The experimental protocol used for the experiments with qualified air traffic controllers is described and our first results are presented. Finally, a global organization involving a multi-level cooperative decision-making is proposed and discussed.