Regulating Trust in Human-Robot Interaction
In collaboration with Matthias Rehm from the University of Aalborg, we investigate what factors contribute to raising trust in interaction, and what factors dampen people's trust in robots. Both can be important: If people trust a robot too little, they will not make use of its capabilities, which can result in their own overload. If they trust robots too much, there can be severe damage if the robot cannot really perform the expected actions. The project is funded by the Danish Council of Independent Research.
In the Nordforsk-funded project Facilitation Robotics for Teams Foundational Research on Group Processes in Human-Robot Ensembles with Social Robots (S-FACTOR), the HRI-Lab Sonderborg is involved with research on how robots can contribute to increasing team creativity.
Together with Kristian Mortensen and Johannes Wagner, Kerstin Fischer organizes a scientific network on Traffic as a Mutually Adaptive System: Interdisciplinary Understanding of the System Characteristics of Traffic. Our partners comprise David Sirkin and Rebecca Currano from the Center for Design Research at Stanford University, Wendy Ju from Cornell Tech, as well as Ole Jensen, Antonia Krummheuer and and Matthias Rehm from Aalborg University.
Two Nordplus Projects on Scandinavian Intonation
Intonation has been found to play a major role in language comprehension. In the following two projects, we aim to enhance the understanding between Scandinavian speakers by giving them the opportunity to familiarize themselves with the prosody and intonation of the other languages.
Development of a Tool for the Visualization of Intonation
In this project, we aim to provide Danish and Swedish speakers with a free tool for training intonation. The tool will visualize target intonation patterns and provide learners with online feedback on the intonation of their own utterances. Besides the speech technology involved, we will develop target descriptions for Danish and Swedish and exercises for students at upper secondary school and university level.
Developing Teaching Material for Speech Prosody to Increase the Mutual Comprehension between Danish, Swedish and Norwegian
In this project, we develop a concept for what teaching materials in prosody for young speakers of Danish, Swedish and Norwegian could look like. We develop exemplary teaching materials for the intonation (speech melodies), stress patterns and other prosodic information that supports speech comprehension. The project aims to promote the Nordic languages and cultures by describing their intonation systems; by presenting an outline for suitable teaching materials; and by publishing scientific articles about the prosodic patterns in these languages and about pedagogical approaches to the teaching of these patterns.
The Trapholt International Museum for Modern Art
Lars Christian Jensen helped the Trapholt museum in Kolding with an exhibition involving our turtlebot.
The Cost of Digital Life – A Citizen Science Project
Socially Intelligent Robotics Consortium
Haru: The experimental social robot from Honda research had been presented at the Socially Intelligent Robotics Consortium.
In the framework of the Socially Intelligent Robotics Consortium, we work on conversational openings and the initiation of interaction, activity-specific use of emotional expression, and role-specific speech and voice characteristics, along with other social cues, to maximize the social affordance of HARU. Furthermore, we study the timing of speech with respect to multi modal social cues and robot persuasiveness and credibility.
The SMOOTH project aims at creating robots for the support of elderly people with a focus on making them interact seamlessly and robustly with their users.
In Sønderborg, we investigate these aspects:
- persuasive dialog
- the effects of responsivity of robot behavior
- the role of incremental speech processing and synthesis
The project is funded by the Innovation Fund Denmark (2017-2020).
The project Improving Second Language Pedagogy at the Prosody-Pragmatics Interface Using Human-Robot Interaction aims to fill the gap regarding the teaching of prosody to second language (L2) learners.
We analyze real conversation with conversation analytic (CA) methods in order to identify the prosodic features that are used by native speakers to fulfil certain communicative tasks. Once we have identified a linguistic feature as a candidate for a trouble source, we use Human-Robot Interaction (HRI) experiments to investigate communicative functions of that feature in a controllable way, by varying one feature at a time in the robots’ behavior, yet still embedding it in an interactional context.
Care-O-bot: Collaboration with DTI and Mærsk-McKinney-Møller Institute
In this collaboration we investigate among other things how a large service robot, such as the Care-O-bot should approach people, pass by people, apply politeness principles and how it can by utilizing several modalities in synchrony afford natural human-robot interactions.
Industrial Robot Teleoperation: Learning by Demonstration
Programming by Demonstration (PbD), is a common way of programming new behaviours into robots and thus addresses “the problems of skill and task transfer from human to robot, as a special way of knowledge transfer between man and machine”. Teaching a robot new behaviours in this way is time effective, intuitive and it can be done by naive users so that costs for expensive programming experts can be saved. In this collaboration we not only investigate the efficiency of different control modalities, but also develop new ones.
Contingency in HRI: Collaboration with Bielefeld & Heriot Watt University
An important aspect of contingency is the temporal proximity between the user’s behavior and the robot’s response. In this collaboration we investigate the degree to which robots that respond in contingent ways are perceived more intelligent, as more competent speakers, and as more engaging. Originally developed for the iCub robot the contingency system was recently ported to work on the EZ-Robot.
Instructing Industrial Robots: Collaboration with University of Innsbruck
How do humans intuitively instruct an industrial robot in a collaborative assembly scenario? How can a robot's gaze be used to ensure smooth and robust human-robot interactions? These and other questions are addressed in our collaboration with the Intelligence and Interactive Systems lab and the University of Innsbruck, a short video of our work is on Facebook.
The Research Network for Transdisciplinary Studies in Social Robotics (TRANSOR) is a platform for research exchange and joint Humanities research in social robotics, connecting researchers in philosophy, robotics, cognitive science, psychology, anthropology, educational science, linguistics, art and design studies, and communication and media studies.
The ITALK project aims to develop artificial embodied agents able to acquire complex behavioural, cognitive, and linguistic skills through individual and social learning. This will be achieved through experiments with the iCub humanoid robot to learn to handle and manipulate objects and tools autonomously, to cooperate and communicate with other robots and humans, and to adapt to changing internal, environmental, and social conditions. More information about the project on the ITALK webpage.