Guy André Boy

CentraleSupélec (Paris Saclay University) & ESTIA, France

Lecture. Introduction to Human-AI Teaming and Human Systems Integration

Abstract.

   Artificial intelligence (AI) has invaded our lives, and human-AI teaming (HAT) should be explored further from a human systems integration (HSI) perspective, more specifically, model-based HSI. HAT is closely associated with the concept of autonomy. The concept of autonomy requires further investigation and a more formal physical and cognitive systemic representation supporting more detailed and meaningful analysis, specifically on situation awareness issues, decision-making, and risk-taking. A new question is: “To what extent should the machine be considered a tool or a partner?”

   Therefore, we need to examine HSI based on a consistent definition of a system where humans and machines can be considered together using a homogeneous representation. More specifically, HAT analysis, design, and evaluation require the development of an appropriate ontology based on HSI and AI principles and concepts, such as HSI-based systems and AI-based agents. Along these lines and beyond the current trend of combining SE and AI, an even stronger need emerges to combine AI and HSI, defined as the intersection of Systems Engineering (SE), Human Factors and Ergonomics (HF/E), and Information Technology (IT).

   AI allows the execution of tasks that usually require human intelligence, such as perception, conversation, and decision-making. AI includes many technologies and applications, such as knowledge-based systems, vision, speech, natural language processing, robotics, machine learning, and planning. If machine autonomy is a contemporary trend, human autonomy remains a significant concern when people must solve problems in unexpected situations using appropriate human skills and knowledge, technologies, and organizational setups. Underlying concepts, methods, and tools are currently being developed by the FlexTech Chair.

   Whenever new life-critical machines, such as manned or unmanned aircraft or spacecraft, are developed, they must be certified to ensure safe, efficient, and comfortable operations. It is, therefore, crucial to develop metrics to evaluate and validate them. HAT-dedicated metrics will be presented from the perspective of machines considered partners, specifically operational performance, trust, and collaboration.

Bio. Guy André Boy, Ph.D., is an INCOSE Fellow, Chair of the HSI Working Group of INCOSE, and Chair of the Aerospace Technical Committee of IEA. He is FlexTech Chair Holder and University Professor at Paris Saclay University (CentraleSupélec) and ESTIA Institute of Technology, France. He was IPA Chief Scientist for Human-Centered Design (HCD) at NASA Kennedy Space Center, Dean of the HCD Institute at the Florida Institute of Technology, USA, President and Chief Scientist of the European Institute for Cognitive Science and Engineering, and Group Lead in the Artificial Intelligence Research Branch at NASA Ames Research Center. He is Fellow of the Air and Space Academy and the International Academy of Astronautics. https://en.wikipedia.org/wiki/Guy_Andr%C3%A9_Boy.

Reading. https://www.sciencedirect.com/science/article/pii/S0160791X23001033