It is striking how visions of the humanoid robot as presented in science fiction utopias and dystopias are driving social robotics research or at least are clearly present in some social robotics research programmes. This is even more surprising since recent successful robotics research took inspiration from biology in the form of insects and other animals. I propose that the history of embodied AI and behavior-based robotics can be regarded a success story of robotics research that brought together researchers from a broad range of disciplines. I will point out some historical highlights and challenges about the collaboration of researchers in computer science, electrical engineering, ethology, biology, philosophy and other fields. This flourishing field in the 1980s led to the advent of commercially successful robots such as autonomous vacuum cleaners and lawn mowers that are now commonplace and often seamlessly interact with humans. Such interdisciplinarity provides a huge challenge for academia, but also for research agencies and administration. Going even beyond academic disciplines, I will also present insights from the more recent interaction of art and robotics and propose elements of a methodology for productive transdisciplinary work in social robotics. I suggest taking inspiration from the animal world rather than views from science fiction to approach the human-robot relation more productively.

The talk will introduce the RDD (research, design, and development process) paradigm of Integrative Social Robotics (ISR;  Seibt 2016), relate it to  “value-sensitive design’”(Friedman, Kahn, Borning 2002)  and “design for values” (van den Hoven 2005) and then discuss whether and in which sense ISR can aspire to generate  a “transdisciplinary research framework” for social robotics applications.  ISR aims to address the  virtually impossible task of developing social robotics applications in a responsible fashion.  Since human socio-cultural interactions are the most complex reality we know, the RDD in social robotics requires maximal interdisciplinary scope; moreover value experience is not only complex, but also highly contextual—stable projections of how people will experience an application are currently not possible.  This creates a gridlock of the “regulation problem” and the “description project” in social robots.  The model of ISR is to address this gridlock, e.g., by intertwining a mixed method approach (i.e., conducting experimental, quantitative, qualitative, and phenomenological research for the same envisaged application) with conceptual and axiological analysis.  We illustrate some of the difficulties of this approach as we currently have encountered them in our research, and in relation to the methodological discussion in HRI research.   In conclusion we consider possible implications of these difficulties for the form of interdisciplinarity of ISR.