Abstract (EN):
Currently, Airline Operations Control Centers (AOCC) are organizations of humans, together with computers to help them in managing the airline operations. In this paper we present MASDIMA a Multi-Agent System that was developed having in mind that the AOCC can be an organization of computers, having humans as helpers in managing the airline operation. MASDIMA has a high degree of autonomy including learning capabilities and is able to integrate information from external systems. MASDIMA learns from data but also from the interaction with the human AOC operators and expects from them feedback and permission before changing the operational plan with the solutions found. MASDIMA will start to be used in a Proof-of-Concept phase in an European Airline in 2016 and preliminary results show that it is able to reduce costs and increase passenger satisfaction. Besides introducing the system we will present our view regarding topics such as the design, analysis and operations of truly integrated systems of humans and machines, based on our experience in designing and operating MASDIMA. Hopefully, our contribution will allow a fruitful discussion upon these important topics. Copyright ISCA.
Idioma:
Inglês
Tipo (Avaliação Docente):
Científica