The aim of this Vinnova-funded project is to develop design guidelines for AI agents to support teamwork. More specifically, the project combines interaction design with ethnomethodology and conversation analysis (EMCA), an approach that analyzes the detailed organization of human interaction. The goal is to iteratively develop a framework for an AI team support system that is adapted to how humans communicate in everyday interaction.
The project will contribute (1) practical knowledge of how ethnomethodology and conversation analysis can be combined with interaction design, (2) a framework for designing AI for collaboration that can be deployed in groups/teams and (3) a data set of groups of people interacting with an autonomous AI agent. The results will be disseminated in an open access paper and through open blogs.
During spring 2022, Hannah Pelikan (Linköping University) will be placed at Cornell University. During this time the group will develop a prototype that is adjusted to human practices for organizing teamwork. We will videorecord and analyze interaction with the prototype and translate knowledge about human interaction patterns into practical design solutions, which will be iteratively evaluated.
Hannah Pelikan 