“Smart” Aircraft: Control in Critical Situations Created by Humans in Flight

Abstract

This article addresses safety in potentially dangerous situations created by negative manifestations of human factors during aircraft flight. These manifestations include erroneous actions by the pilot, delayed reaction to rapid changes in flight conditions, inattention, fatigue, illness, inaction, suicidal intent, hijacking of the aircraft by intruders, including terrorists, who are among the passengers, panic behavior of passengers, etc. The escalation of such a situation, considered critical, into an accident can be prevented if the aircraft is designed as a “smart” human-machine system with a high level of robotization. The automatic part of this system must be able to recognize dangerous human behavior and perform autonomous measures aimed at minimizing the risks. In the most extreme case, it must block human actions and transfer aircraft control to a fully automatic mode – until the end of the flight with a safe landing. The purpose of the article is to formulate the problem of developing an onboard automatic control system that meets such tasks, and a preliminary analysis of the possibilities of its solution. The specific features of the problem under consideration determine the choice of artificial intelligence elements, in particular neural network technology, for its effective solution.