I bring you a weekly bite-sized chunk of the science behind helicopter human factors and CRM in practice, simplifying the complex and distilling a helicopter related study into a summary of less than 500 words.
TITLE:
Human Factors requirements for Human-AI teaming in aviation.
WHAT?
Paper proposing a comprehensive Human Factors framework for Human-AI Teaming (HAT) in aviation. Recognising that AI-enabled intelligent agents will increasingly support pilots and air traffic controllers, the study considers what would be needed to ensure that such partnerships remain safe, effective and human-centred.
WHERE?
Conducted by Barry Kirwan at EUROCONTROL, it formed part of the HAIKU project (Human–AI Teaming Knowledge and Understanding for Aviation). This is a research programme involving a consortium of European universities, research institutes, air navigation service providers, and aerospace companies gathered to develop and evaluate safe, human-centred forms of Human-AI Teaming (HAT).
WHEN?
Published in Future Transportation, April 2025.
WHY?
AI promises major safety and efficiency benefits, but its limitations, such as opacity, bias and unexpected behaviour, create new risks. Future intelligent agents will fundamentally alter pilot-automation and controller-automation relationships, requiring an evolution of Human Factors principles rather than incremental adaptation.
HOW?
The study examined three prototype AI systems to build a framework of 8 HF domains and 165 requirements to ensure the safe and effective development of AI from a human factors perspective. They included areas such as human-centred design; Human-AI interaction; trust and explainability; situation awareness and workload; training and competence; teaming and resilience; and organisational readiness.
It used three prototype systems as test cases:
FOCUS – a cockpit assistant for single-pilot operations. If a pilot experiences a startle, the AI detects physiological signs of stress and highlights the most important instruments to help the pilot regain situational awareness.
COMBI – an intelligent diversion assistant to helps crews identify suitable alternate airports during severe weather or airport closures.
ISA (Intelligent Sequence Assistant) – helps controllers optimise arrival and departure sequencing while providing explanations for its recommendations.
FINDINGS:
Human Factors frameworks developed for conventional automation are not fully adequate for Human-AI Teaming. Several long-established concepts such as Human-Centred Design, Crew Resource Management, situation awareness, workload management and resilience engineering need revisiting to address AI-specific issues such as trust calibration, operational explainability and shared autonomy.
Application of the 165 requirements to the prototype systems generated new design insights and demonstrated that the framework is scalable across different operational domains and levels of AI autonomy.
SO WHAT?
This paper provides one of the earliest systematic attempts to define Human Factors requirements for operational Human-AI Teaming.
Its central message is that AI should not be viewed simply as more sophisticated automation. Once intelligent agents begin sharing tasks, initiating actions and negotiating solutions, traditional assumptions underlying automation may no longer hold.
It suggests that future CRM, training systems, certification processes and SMS frameworks will need to accommodate AI as a collaborative team member rather than as a tool. Operational explainability, trust management and maintaining meaningful human authority are likely to become core Human Factors challenges of the next decade.
REFERENCE:
Kirwan, B. (2025). Human factors requirements for human-AI teaming in aviation. Future Transportation, 5(2), Article 42. https://doi.org/10.3390/futuretransp5020042
competentaviators.com 