What if artificial intelligence piloted planes? Discover the secrets of the future of flight!

Imagine a future where artificial intelligence (AI) collaborates with pilots to make flights safer. Within the next decade, AI could revolutionize aviation, but not without challenges. Current AI systems have limitations, such as data biases and 'hallucinations', so it's essential that humans monitor and integrate AI decisions. The European Union has already established that in critical sectors like aviation, AI must be supervised by humans. However, policies alone are not enough: methods and validations are needed to ensure effective human-AI collaboration. This study explores the different types of AI and levels of autonomy considered for future aviation. It also identifies key areas where human factors can improve the safety and performance of human-AI teams. The result? A set of 165 detailed requirements, already tested on AI prototypes in cockpits and air traffic control centers. These requirements could be the key to a future where humans and AI work together seamlessly.

The study delves into eight key areas of Human Factors for human-AI teaming: Human-Centered Design, Human-Machine Interaction, Communication, Collaboration, Reliability, Error Management, Training and Preparedness, and Organizational Readiness. Each area was analyzed to generate specific requirements that ensure effective collaboration. For example, Human-Centered Design requires that AI interfaces be intuitive and easy to use, while Error Management focuses on how AI can help prevent and manage human errors. The requirements were tested on three AI-based intelligent agent prototypes, two for the cockpit and one for air traffic control. The results suggest that these requirements are scalable and adaptable to different levels of design maturity and AI autonomy.