Dear Colleagues,

The use of AI and ML for Autonomous Aircraft became an indispensable element to meet the new aviation mission requirements, either in civil or military applications. However, guaranteeing safety and security of the ML / AI modules is a considerable challenge to be solved in the decades to come. Industry and regulatory bodies are proposing a phased approach, whereby the assurance efforts will be focused on progressive levels of application complexity, starting by Human Assistance (Level 1), advancing to Human-Machine Collaboration (Level 2), and then to Highly Autonomous Systems (Level 3). A starting chapter in this roadmap is the Trustworthiness Analysis, in which the main ethical guidelines are established by representative bodies of the broad society, with characterization of the AI/ML applications, followed by the initial safety and security assessments. The next pre-requisites towards acceptability are the development of AI Assurance means, analysis of the relationship between Human Factors and AI, and construction of Risk Mitigation Means. These are the building blocks of a road to certification. The presence or not of explainability is a key property in deciding whether or not some technology meets the acceptance criteria.

We consider an autonomous aircraft as a system within a larger System-of-Systems constituted by the airspace and other aviation participants, including embedded, airborne systems, and ground-based control stations and Air Traffic Management Systems. Because of the stringent weight and power consumption requirements for the airborne systems, it is expected that a large portion of the mission management intelligence and databases will remain with ground systems, thus we are dealing with a highly distributed system, that must be resilient to communication failures.

Within this context, all methods and techniques in AI/ML may find use cases where they offer differential benefits, thus worth of research efforts and publication in this section. These benefits can be better performance, higher efficiency and increased safety, just to name a few that can be subject of studies in this section.

We look forward to receiving your contributions.

Dr. Ítalo Romani de Oliveira

Section Editor

Autonomous Aircraft; ML / AI modules; Human-Machine Collaboration; Highly Autonomous Systems; Human Assistance; Risk Mitigation Means; Air Traffic Management Systems; AI Assurance Means