Research

Multimodal foundation models empower robots with unprecedented zero-shot scene understanding. We leverage, for example, Vision-Language Models (VLMs) for open-set object recognition and scene comprehension, enabling downstream tasks such as semantic modeling and mobile manipulation. Our research explores areas like open-vocabulary semantic SLAM in labeling-scarce environments and learning intra- and inter-object 3D scene graphs for affordance grounding and manipulation.

Selected Publications:

• FunGraph: Functionality Aware 3D Scene Graphs for Language-Prompted Scene Interaction, D. Rotondi, F. Scaparro, H. Blum, K.O. Arras, IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS), 2025.
• Long-Term Planning Around Humans in Domestic Environments with 3D Scene Graphs, E. Bartoli, D. Rotondi, K.O. Arras, I. Leite, Workshop at the ACM/IEEE Int. Conf. on Human-Robot Interaction (HRI), 2025
• NounAtlas: Filling the Gap in Nominal Semantic Role Labeling, R. Navigli, M. Pinto, P. Silvestri, D. Rotondi, S. Ciciliano, A. Scirè, Annual Meeting of the Association for Computational Linguistics (ACL), vol 1, 2024 (Outstanding Paper Award).
• Plug-and-play SLAM: A unified SLAM architecture for modularity and ease of use, M. Colosi, I. Aloise, T. Guadagnino, D. Schlegel, B. Della Corte, K.O. Arras, G. Grisetti, IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS), 2020.

Robots operating in human environments must possess social skills, from navigating crowds in socially normative ways to using natural and engaging non-verbal cues. We are particularly interested in social capabilities to engage users in the long-term, after novelty effects have worn off, and ask: what makes interactive robots sustainably believable and effective? Our research explores, for example, generative AI for robot design or learning techniques and representations for automatically generating rich, expressive, non-verbal motions across diverse robot morphologies.

Selected Publications:

• Interactive Expressive Motion Generation Using Dynamic Movement Primitives, T. Hielscher, A. Bulling, K.O. Arras, IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS), 2025.
• Incremental and Interactive Exploration of Robot Appearance Designs Using GenAI, T. Hielscher, A. Bartenbach, W. Leonhardt, K.O. Arras, IEEE Int. Conf. on Robot and Human Interactive Communication (RO-MAN), 2025.
• The Child Factor in Child-Robot Interaction: Discovering the Impact of Developmental Stage and Individual Characteristics, I. Rudenko, A. Rudenko, A.J. Lilienthal, K.O. Arras, B. Bruno, Int. Journal of Social Robotics, 16(8), 2024.
• SPENCER: A Socially Aware Service Robot for Passenger Guidance and Help in Busy Airports, R. Triebel, K.O. Arras et al., In Field and Service Robotics, Springer Tracts in Advanced Robotics, vol 113, 2016. EU Project Link.

Rapid advancements in robotics and AI for embodied intelligence are unlocking new levels of flexibility, dexterity, and autonomy for complex tasks in unstructured environments. Our research integrates model-based approaches with data-driven robot learning techniques to meet the demands of real-world applications. Examples include bin picking and bin packing, or industry-savvy SLAM solutions. With extensive expertise in robotics technology transfer, we collaborate with the Fraunhofer Society and industry leaders to drive innovation in AI-powered robotics.

Selected Publications:

• Snuggle-Pack: Speeding Up Multi-Heuristic Packing Planning of Complex Objects, T. Nickel, R. Bormann, K.O. Arras, IEEE/RSJ Int. Conf. on Intelligent Robots and Systems (IROS), 2025
• Multi-Heuristic Robotic Bin Packing of Regular and Irregular Objects, T. Nickel, R. Bormann, K.O. Arras, IEEE Int. Conf. on Robotics and Automation (ICRA), 2025
• The e-Bike Motor Assembly: Towards Advanced Robotic Manipulation for Flexible Manufacturing, L. Rozo, A. Kupcsik, P. Schillinger, M. Guo, R. Krug, N. van Duijkeren, M. Spies, P. Kesper, S. Hoppe, H. Ziesche, M. Bürger, K.O. Arras, Robotics and Computer-Integrated Manufacturing, vol. 85, 2024
• The ILIAD Safety Stack: Human-Aware Infrastructure-Free Navigation of Industrial Mobile Robots, S. Molina, A. Mannucci, M. Magnusson, D. Adolfsson, H. Andreasson, M. Hamad, S. Abdolshah, R. Chadalavada, L. Palmieri, T. Linder, C. Swaminathan, T. Kucner, M. Hanheide, M. Fernandez-Carmona, G. Cielniak, T. Duckett, F. Pecora, S. Bokesand, K.O. Arras S. Haddadin, A.J. Lilienthal, IEEE Robotics & Automation Magazine, 2023. EU Project Link.