FOCUS ARTICLE Robotic interfaces for cognitive psychology and embodiment research: A research roadmap Philipp Beckerle 1,2 | Claudio Castellini 3 | Bigna Lenggenhager 4 1 Elastic Lightweight Robotics Group, Robotics Research Institute, Technische Universität Dortmund, Dortmund, Germany 2 Institute for Mechatronic Systems in Mechanical Engineering, Technische Universität Darmstadt, Darmstadt, Germany 3 Institut of Robotics and Mechatronics, DLR German Aerospace Center, Oberpfaffenhofen, Germany 4 Cognitive Neuropsychology, Department of Psychology, University of Zurich, Zurich, Switzerland Correspondence Philipp Beckerle, Elastic Lightweight Robotics Group, Robotics Research Institute, Technische Universität Dortmund, Dortmund, Germany. Email: philipp.beckerle@tu-dortmund.de Funding information Swiss National Science Foundation; German Research Foundation, Grant/Award Numbers: CA 1389/1, BE 5729/3 and BE 5729/11 Advanced human–machine interfaces render robotic devices applicable to study and enhance human cognition. This turns robots into formidable neuroscientific tools to study processes such as the adaptation between a human operator and the operated robotic device and how this adaptation modulates human embodiment and embodied cognition. We analyze bidirectional human–machine interface (bHMI) technologies for transparent information transfer between a human and a robot via efferent and afferent channels. Even if such interfaces have a tremendous positive impact on feedback loops and embodiment, advanced bHMIs face immense technological challenges. We critically discuss existing technical approaches, mainly focusing on haptics, and suggest extensions thereof, which include other aspects of touch. Moreover, we point out other potential constraints such as limited functionality, semi-autonomy, intent-detection, and feedback methods. From this, we develop a research roadmap to guide understanding and development of bidirectional human–machine interfaces that enable robotic experi- ments to empirically study the human mind and embodiment. We conclude the integration of dexterous control and multisensory feedback to be a promising road- map towards future robotic interfaces, especially regarding applications in the cog- nitive sciences. This article is categorized under: Computer Science > Robotics Psychology > Motor Skill and Performance Neuroscience > Plasticity KEYWORDS cognitive science, embodiment, human–machine interfaces, robotics, sensory substitution 1 | INTRODUCTION Human–machine interfaces (HMIs) are advancing so rapidly that they almost unexpectedly developed into a formidable tool to study the human mind and its underlying neural mechanisms. Recent HMIs provide a unique possibility to study adaptive processes between human beings, robots, and the environment, which is interesting for a variety of theoretical and applied fields; most excitingly may be the study of embodiment and embodied cognition and with it the plasticity of the bodily self (Apps & Tsakiris, 2014; Blanke, 2012). Studying to what extent and under which preconditions humans adapt to a machine might inform us about basic mechanisms of embodiment and embodied cognition, which are typically difficult concepts to study since the human body is “always there” (James, 1890) and cannot easily be experimentally modified (Tsakiris & Haggard, 2005). Received: 25 May 2018 Revised: 3 October 2018 Accepted: 20 October 2018 DOI: 10.1002/wcs.1486 WIREs Cogn Sci. 2018;e1486. wires.wiley.com/cogsci © 2018 Wiley Periodicals, Inc. 1 of 9 https://doi.org/10.1002/wcs.1486