Descripción del proyecto
Estudio de los sistemas inteligentes de conciencia situacional del espacio para las operaciones de control de tráfico aéreo
La automatización constituye una solución prometedora para el problema de capacidad en la gestión del tráfico aéreo. Sin embargo, para aplicar conceptos de automatización avanzada, los humanos y los sistemas de IA tienen que compartir la conciencia situacional del espacio. Por consiguiente, el proyecto financiado con fondos europeos AISA tiene como objetivo investigar el efecto de la conciencia situacional del espacio distribuida humano-máquina en operaciones de control del tráfico aéreo en ruta, así como estudiar las oportunidades que entraña. Para ello, el proyecto no se centrará en la automatización de tareas aisladas, sino que desarrollará un sistema inteligente con conciencia situacional del espacio. Dicho sistema artificial allanará el camino a la automatización avanzada del futuro basada en el aprendizaje automático.
Objetivo
This proposal addresses the topic “Digitalisation and Automation principles for ATM”. Automation is one of the most promising solutions for the capacity problem, however, to implement advanced automation concepts it is required that the AI and human are able to share the situational awareness. Exploring the effect of, and opportunities for, distributed human-machine situational awareness in en-route ATC operations is one of the main objectives of this project. Instead of automating isolated individual tasks, such as conflict detection or coordination, we propose building a foundation for automation by developing an intelligent situationally-aware system. Sharing the same team situational awareness among ATCO team members and AI will enable the automated system to reach the same conclusions as ATCOs when confronted with the same problem and to be able to explain the reasoning behind those conclusions. The challenges of transparency and generalization will be solved by combining machine learning with reasoning engine (including domain-specific knowledge graphs) in a way that emphasizes their advantages. Machine learning will be used for prediction, estimation and filtering at the level of individual probabilistic events, an area where it has so far shown great prowess, whereas reasoning engine will be used to represent knowledge and draw conclusions based on all the available data and explain the reasoning behind those conclusions. We will explore to what extent it is possible to deduce machine learning false estimates and how resilient such system is to failure. In this way, the artificial situational awareness system will be the enabler of future advanced automation based on machine learning.
Ámbito científico
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringcontrol systems
- social sciencessociologyindustrial relationsautomation
- natural sciencescomputer and information sciencesknowledge engineering
- natural sciencescomputer and information sciencesartificial intelligencemachine learning
Palabras clave
Programa(s)
Régimen de financiación
RIA - Research and Innovation actionCoordinador
10000 Zagreb
Croacia