Research Methodologies in Human Robot Interaction
Research in human robot interaction is an inherently interdisciplinary field. Besides the technical aspects of engineering research, artificial intelligence and machine learning approaches meet linguistic models, psychologically motivated behaviors, as well as usability aspects of interface design. While it is fairly straight forward to implement an HRI system that works once in the lab for a specific scenario demonstrated by the developer team, any further generalization or deeper analysis - in any of the research directions mentioned before - requires more sophisticated methodologies. At Bielefeld University with its interdisciplinary research center on cognitive interaction technology (CITEC) this is a cross-cutting direction of research that has lead to a significant effort for improved HRI experimentation. In my talk, I will touch several problems and challenges related to an interdisciplinary view of human-robot interaction research. Taking different perspectives, I will discuss issues of human expectations and influencing factors in HRI experiments, issues of system complexity versus the requirement of experimental reproducibility, as well as issues of ecological validity versus experimental simplicity. Examples shown in the talk range from behavioral experiments to human-robot handovers and verbal as well as augmented interaction with robot companions in domestic environments.
Short bio:
Sven Wachsmuth received the Diploma and Ph.D. degree in computer science from Bielefeld University, Germany, in 1997 and 2001. He did his habilitation in 2009 in the area of applied computer science. In 1997 he joined the Applied Informatics Group (Angewandte Informatik) at the Faculty of Technology, Bielefeld University. From 1997 to 2000 he worked within the collaborative research centre 360 "Situated Artificial Communicators". Since 2001, he holds a faculty staff position at Bielefeld University, first as a senior researcher (akad. Rat) in the Applied Computer Science Group, then (since 2009) as a lecturer (akad. Oberrat). He was announced as the head of the central lab facilities of the Excellence Cluster 277 "Cognitive Interaction Technology" in 2008. In Nov 2002, he took a sabbatical year supported by the German Research Foundation (DFG) at the Computer Science Department of the University of Toronto as a host of Prof. Sven Dickinson. Between 2006 and 2018, he was principle investigator in several research projects such as the special research centre 673 "Alignment in Communication" (2006-2014; A4, C1), the Excellence Cluster 277 "Cognitive Interaction Technology" (2007-2018), BMBF project KogniHome (2014-17), and the BMBF project ADAMAAS (2015-18). His research interests include human-robot interaction, scene understanding for robot vision, linking computer vision and language, service robotics, cognitive assistance as well as system integration and evaluation. In the last years, a special emphasis has been on reproducible experimentation in robotics. Since 2013, he has been a member of the RoboCup@Home executive committee and has been the local chair of the RoboCup@Home league 2016 in Leipzig.