Non Humanoid Robot
Non-humanoid robots encompass a diverse range of designs, from multi-legged robots inspired by insects to wheeled and even flapping-wing robots, all aiming to achieve efficient and adaptable locomotion and manipulation in various environments. Current research emphasizes improving robot autonomy through reinforcement learning, particularly for gait generation and task planning, often incorporating large language models (LLMs) for natural language instruction processing and human-robot interaction. These advancements are significant for expanding robotic capabilities in challenging tasks such as cooking, search and rescue, and collaborative assembly, ultimately impacting fields ranging from manufacturing and healthcare to exploration and disaster response.
Papers
RAPTOR: Rapid Aerial Pickup and Transport of Objects by Robots
Aurel Appius, Erik Bauer, Marc Blöchlinger, Aashi Kalra, Robin Oberson, Arman Raayatsanati, Pascal Strauch, Sarath Suresh, Marco von Salis, Robert K. Katzschmann
Interactive Disambiguation for Behavior Tree Execution
Matteo Iovino, Fethiye Irmak Doğan, Iolanda Leite, Christian Smith
Self-Supervised Learning for Joint Pushing and Grasping Policies in Highly Cluttered Environments
Yongliang Wang, Kamal Mokhtar, Cock Heemskerk, Hamidreza Kasaei
Teaching Robots to Span the Space of Functional Expressive Motion
Arjun Sripathy, Andreea Bobu, Zhongyu Li, Koushil Sreenath, Daniel S. Brown, Anca D. Dragan