‘Yes dear, that belongs into the shelf!’ - Exploratory studies with elderly people who learn to train an adaptive robot companion

2Citations
Citations of this article
10Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Robot companions should be able to perform a variety of different tasks and to adapt to the user’s needs as well as to changing circumstances. To achieve this we can either built fully adaptive robots or adaptable and customizable robots. In this paper we present an adaptable companion which uses a decision making algorithm and user feedback to learn adequate behavior in new tasks. Using two different scenarios (household task, card game) the system was evaluated with elderly people in exploratory studies. We found that the perception and evaluation of the robot’s learning progress depends on the interaction scenario. Additionally, we discuss improvements for the algorithm in order to make the learning behavior appear more natural and humanlike.

Cite

CITATION STYLE

APA

Hoefinghoff, J., Rosenthal-von der Pütten, A., Pauli, J., & Krämer, N. (2015). ‘Yes dear, that belongs into the shelf!’ - Exploratory studies with elderly people who learn to train an adaptive robot companion. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9388 LNCS, pp. 235–244). Springer Verlag. https://doi.org/10.1007/978-3-319-25554-5_24

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free