An Ontological Approach to Autonomous Navigational Decision Making in Aircraft Systems
Paul Vajda · Spectrum Research Repository (Concordia University) · 2020
Aircraft systems are becoming increasingly complex, and automation increasingly necessary for safe aircraft operation. Current industry trends are encouraging either a reduction of crew num- bers or complete elimination. After surveying the evolution of aircraft automation systems, pilot tasks and situational awareness, we argue that a hurdle to crew reduction is that of autonomous navigation. In this thesis, we put forth an ontology based approach for defining autonomous navigational decision making. The ontology is designed to capture elements representing the environment in which aircraft operate; the “air environment”. An expert system is then pro- grammed and uses the ontology as its knowledge base. Combining the environmental ontology with a second one describing a generic air vehicle, the expert system is programmed to make navigational decisions based on environment, vehicle state and mission. Case studies show the functionality of the implementation, and a validation study demonstrates the feasibility of the solution. The thesis closes by relating it to other research and recently released industry avionics solutions.