go back

Predictive and Robust Robot Assistance for Sequential Manipulation Tasks

Theodoros Stouraitis and Michael Gienger, "Predictive and Robust Robot Assistance for Sequential Manipulation Tasks", IEEE Research and Automation Letters (RA-L), 2023.


This paper presents a novel concept to support impaired users in daily physical object manipulation tasks with a robot. Starting out with an assumed manipulation task of a user, we propose a predictive model that uniquely casts the user's sequential behavior as well as a robot support intervention into a hierarchical multi-objective optimization problem. A major contribution is the prediction formulation, which allows to model several different future paths concurrently. The second contribution is the encoding of a general notion of constancy constraints which allows to consider long-term dependencies between the keyframes of a longer task. We evaluate the concepts in a table top task where a robot supports impaired users by proactively re-arranging objects.

Download Bibtex file Download PDF