Artificial Intelligence and Relations to Avionics

In this course we are going to introduce concepts of decision making conducted by algorithms which led to current term Artificial Intelligence. The course is built around flight planning algorithms, their performance and suitability for different applications. Within our session we will focus on and summarize advantages and disadvantages of Breadth First Search, A*, Iterative Deepening A*, Theta*, and RRT* algorithms. Their reasoning process and path selection methodology with perspective of aerospace requirements are evaluated. Our focus will be on the randomization element and uncertainty of these algorithms. We will also describe selected evaluation parameters required by FAA and EASA Technical Standard Order (TSO) documents on electronic systems and what are the conflicts between these requirements and the natural principle of the existing path-planning algorithms. The influence of the performance of the navigation sensors and expected departure and arrival procedures which use the existing navigation means (INS, VOR, NDB, ILS, GPS) will be discussed. Finally, we describe the Artificial Intelligence phenomena and discuss the determinism of the currently used algorithms for flight-path panning and recovery.

Dr. Pavel Paces works at Artificial Intelligence Center at Faculty of Electrical Engineering, Czech Technical University in Prague, Czech Republic. He gained MSc and Ph.D. in aerospace engineering. Currently, he leads a group developing solutions for pilot training evaluation and fatigue measurement aiming on human machine interaction, processes and procedures. He received Honeywell Innovator award in 2011, he is member of IEEE Aerospace and Electronic Systems Society and AIAA.