Assistant Professor · Human Media Interaction · University of Twente
I study how robots can coach people to exercise and stay healthy, and how they can adapt to each person they work with.
s.schneider@utwente.nl University profile Google Scholar ORCID
I am an Assistant Professor in the Human Media Interaction group at the University of Twente. My research is on AI and robotics for health coaching and on social-physical human-robot interaction: robots that exercise with people, give feedback, and learn each person’s preferences through interaction.
I received my PhD in Intelligent Systems from Bielefeld University in 2019, with a thesis on socially assistive robots for exercising. Since then I have worked on persuasive robots in public spaces at Kyoto University, on text mining for an ERC project in sociology, on conversational AI and virtual agents for train stations and museums at TH Köln, and as a data scientist building machine-learning products at Miele.
At Twente I teach Human-Robot Communication and Human-Robot Collaboration and co-supervise three PhD candidates.
Two papers presented at IEEE RO-MAN 2026 in Kitakyushu
“ISCA: A Structured Framework for Assessing and Reporting of Interactive System Complexity”, with Kira Loos and Anna-Lisa Vollmer, and “Nonvisual Human–Robot Handovers: Multimodal Design and Evaluation”, with Arthur van der Torre, Daan Doeleman and Casper ten Holder. The proceedings are still to appear.
Bas van der Kaaden and Lydia Penkert started as PhD candidates.
I will be the daily supervisor for two PhD candidates. Bas (Utwnete) works on social-physical human-robot interaction for physiotherapy and coaching, Lydia (Bielefeld University) on the impact of Robotic Social touch on neurotypical and cognitively impaired patient experiences in healthcare.
New paper at HRI 2026 in Edinburgh
“Learning from Human-Dog Collaboration: Towards the Design of Search and Rescue Robots”, with Katharina Lisa Kleiser and Veerle Buntsma.
Socially assistive robots that accompany people through indoor cycling, rowing, squats and rehabilitation exercises, and the feedback and motivation strategies that keep people going over weeks rather than minutes.
See: Exercising with a humanoid companion is more effective than exercising alone (Humanoids 2016)
Preference learning, dueling bandits and reinforcement learning that let a robot adapt its coaching to one individual, and studies of how people teach robots and how they judge a robot that adapts on its own.
See: Exploring embodiment and dueling bandit learning for preference adaptation (RO-MAN 2017)
What happens when robots and virtual agents leave the lab for train stations, museums and shopping malls: being noticed, being taken seriously, and persuading people to follow rules.
See: Stop ignoring me! On fighting the trivialization of social robots in public spaces (ACM THRI 2022)
Touch, handover and physical guidance between people and robots, including for users who are blind or visually impaired, people with cognitive impairement, and patients in physiotherapy.
See: Social-physical human-robot interaction for health applications: A scoping review (ICSR 2024)
Robots that coach through touch and physical guidance as well as speech, for physiotherapy and exercise. Recent work combines haptic feedback with a socially assistive robot to improve squat technique.
Designing robot interaction with and for people who are often left out of studies: object handover under reduced vision, sound-based guidance for blind users, and social touch with people with intellectual disabilities.
Agents that learn to work together from intrinsic motivation instead of hand-designed rewards, studied in cooperative multi-agent environments such as Overcooked.
Funding: this work is supported by the starting package of my assistant professorship, financed through the Dutch Sectorplan Bèta en Techniek.
Citations and newer papers: Google Scholar · ORCID
I supervise BSc and MSc theses, internships and research projects at the University of Twente, for students from programmes such as Creative Technology, Technical Computer Science, Interaction Technology and Robotics.
If one of these areas fits your interests, send an email to s.schneider@utwente.nl with your study programme, the kind of project (thesis, internship, Research Topics or Capita Selecta) and your planned start date. A short paragraph on what you would like to work on is enough.
Examples of completed theses are listed under Teaching.
I currently have no funding for PhD positions. If you have your own funding, for example a scholarship or fellowship, I am happy to welcome you and to talk about a project.
Guest researchers and visiting students are welcome as well.
Assistant Professor NowHuman Media Interaction, University of Twente
Research on AI and robotics for health coaching and social-physical human-robot interaction. Designs and teaches the human-robot interaction courses of the MSc Robotics.
Data ScientistMachine Learning and AI Group, Miele & Cie. KG
Data-driven prototypes, services and products for connected appliances, from model design to rollout and monitoring (MLOps with mlflow, dvc and Azure ML).
PostdocCologne Cobots Lab, TH Köln
Coordinated the BMBF-funded project SKILLLED with DB Systel: conversational AI, virtual agents and robots for public spaces such as train stations and museums.
PostdocFaculty of Sociology, Bielefeld University
Text mining and topic modeling on the rise of precision medicine over two decades, in the ERC-funded Predict project.
Specially appointed researcherHRI Laboratory, Kyoto University
JST-funded work on persuasive mobile robots that increase rule adherence in public spaces, and on how people perceive a robot’s authority and morality.
Research associateApplied Informatics, Bielefeld University
Lectured Social Robotics; preference-based reinforcement learning for robot skill learning; human-robot handover under visual impairment.
Research scholarCITEC Center of Excellence, Bielefeld University
Robotic exercising companions for health care, and a general framework for building socially assistive interaction scenarios.
Research scholarApplied Informatics, Bielefeld University
Taught social robotics; analysis and dissemination of long-term human-robot interaction data.
Research associateCognition and Robotics Lab, Bielefeld University
Long-term motivational strategies for socially assistive robots in the SoziRob project, with longitudinal experiments run together with the German Aerospace Center (DLR).
Human Media Interaction, Faculty of Electrical Engineering, Mathematics and Computer Science, University of Twente.
Citadel, room H208, Hallenweg 15, 7522 NH Enschede, Netherlands.
Sebastian Schneider is an Assistant Professor in the Human Media Interaction group at the University of Twente. His research focuses on socially assistive robots for health coaching, social-physical human-robot interaction, and interactive robot learning that adapts to individual users. He received his PhD in Intelligent Systems from Bielefeld University in 2019 and has held research positions at Kyoto University, Osaka University, Bielefeld University and TH Köln, as well as a data science role at Miele. He has co-organized workshops on personal robots for exercising and coaching and on human-robot interaction in public spaces at the ACM/IEEE HRI conference, and is Registration Co-Chair of HRI 2027.
For talk announcements and programs.