AIR - Action and Intention Recognition in Human Interaction with Autonomous Systems

AIR investigates action and intention recognition in human interaction with autonomous systems (or AIR, for short). More specifically, the focus is on the interaction of humans and autonomous systems that move in shared physical spaces.

The mutual recognition of actions and intentions between humans and the autonomous systems they interact with is absolutely crucial to ensure safety as well public acceptance of such technologies. Therefore, the research to be performed within the AIR distributed research environment will contribute to the success of future autonomous systems in society.

The particular research scenarios that instantiate this general theme in the AIR environment address human interaction (1) with social/assistive robots in the home, (2) with autonomous transport vehicles in industrial environments, and (3) with autonomous vehicles in public traffic.

All of these scenarios share the requirements that the autonomous systems must behave in a way that is unobtrusive and transparent to the humans interacting with them, and have cognitive abilities that allow humans to intuitively and effortlessly communicate their intentions and desired actions. Each scenario, however, considers different facets of the problem.

Addressing the challenge of mutual action and intention recognition therefore requires a well-coordinated long-term interdisciplinary research effort that integrates expertise in cognitive science, experimental psychology, human factors, human-machine interaction, robotics, computer vision, and artificial intelligence - as constituted by the AIR consortium.