HAIKU: Human AI teaming Knowledge and Understanding for aviation safety

HAIKU aims to deliver and study mature prototypes of AI Digital Assistants for the aviation domain to derive comprehensive human factors design guidance and methods on how to develop safe, effective, and trustworthy Digital Assistants for Aviation. Moreover, HAIKU will output new safety and validation assurance methods for Digital Assistants, to facilitate early integration into aviation systems by aviation stakeholders and regulatory authorities. HAIKU will develop Digital Assistants in six different use cases:

• Digital Assistant for Urban Air Mobility (UAM) to assist UAM coordinators in managing repetitive tasks, traffic monitoring, and dealing with non-standard situation during planning and operations. This use case is lead by LiU with support from LFV, and builds on existing technology developed at LiU including UTM City and SOMA-AI.
• Digital Assistant in the flightdeck to assist pilots in ‘startle response’ adverse events.
• Digital Assistant in the flightdeck to assist pilots in route planning/replanning.
• Digital Assistant in tower (and remote tower) to assist controllers in routine and repetitive tasks for aircraft on approach.
• Digital Assistant for Urban Air Mobility (UAM) to assist UAM coordinators in managing repetitive tasks, traffic monitoring, and dealing with non-standard situation during planning and operations. This use case is lead by LiU with support from LFV, and builds on existing technology developed at LiU including UTM City and SOMA-AI.
• Digital Assistant in airports to assist safety management in data collection, categorization, analysis, and visualisation.
• Digital Assistant in airports to assist in monitoring risk factor conditions of infectious disease (e.g. COVDI-19) and prevention by providing information to passengers on safe paths.

HAIKU will guide the development of human-centric-AI in the aviation industry by developing guidance and assurance procedures, and by exploring Human-AI Teaming via a number of interactive prototypes. The main research questions are:

1. What is the recommended human-AI relationship for each of the different AI applications in aviation?
2. What does it mean for AI to be explainable and hence trustworthy in each of these applications?
3. How do we best teach AI, via human-in-the-loop AI learning for each of the potential aviation applications?

HAIKU is an innovative actions research project receiving funding from the Horizon 2020 research and innovation program. The project runs for three years between 2022-2025 and comprise 18 international partners from 12 countries.