This special session on Cognition for Human-Robot Interaction will focus on the use of cognitive architectures for social robots in human-machine interaction. If humans and machines are to work together as partners, they must have the same basic understanding of a task and the current situation. Conclusions of mental models from cognitive architectures might be the basis for the actions of a social robot. Different methods and integration of methods can be used as a kind of control instance for its sensors and, for example, the utterances generated via an LLM. Cognitive principles can help to make the robot's behavior more understandable, comprehensible and reliable for humans. Integrating such cognitive aspects could improve human-robot interaction and increase trust, acceptance and mutual understanding.
Relevant questions:
- Which cognitive architectures are suitable for social robots and how can they be implemented?
- In what way are cognitive principles relevant for human-robot interaction?
- What possibilities are there for learning individual preferences and behaviors of the human interaction partner?
- In what way does belief formation and updating support better and more fluent interaction?
- What technical possibilities are there to increase trust, acceptance and mutual understanding and what opportunities do they offer?
List of topics:
- Cognitive architectures for human-machine interaction
- Cognitive models for human-machine interaction
- Cognitive models for human-machine collaboration
- Cognitive modeling & knowledge representation for robots
- Generative AI for situation awareness and situated human-machine interaction
- Situation sensing, perception, comprehension, and prediction in human-machine interaction
- Theory of mind in human-machine interaction
- Mental models & mental states
- Belief updating
- Systems, platforms and tools for cognition for human-machine interaction
- Technical implementations
