ROBOTIS (Robot is ...)

Tommoro Robotics is pleased to announce the signing of an MOU with ROBOTIS (Robot is ...), marking the beginning of a strategic partnership to accelerate the commercialization of Physical AI humanoid robots. As part of this collaboration, Tommoro Robotics has recently purchased an additional 5 units of ROBOTIS’s new “AI Worker” platform. Building on this, we will work closely with ROBOTIS to equip these humanoids with our VLA-based Robot Foundation Model (RFM), aiming for true commercial deployment. By combining ROBOTIS’s high-precision hardware—such as DYNAMIXEL actuators and DYD reducers—with Tommoro’s unique RFM and advanced AI-driven control software, we are developing next-generation humanoids that can understand and execute complex tasks. Together, our goal is to make humanoid robots that are truly deployable across logistics and a variety of industrial environments, while also contributing actively to the academic and open-source communities. This partnership is more than a technical alliance; it’s a joint commitment to realize the commercial potential of Physical AI Agents that can genuinely collaborate with people in real-world settings. With ROBOTIS, we are ready to usher in a new era of human-centric, AI-powered humanoid robotics! https://lnkd.in/gMcFSmDD

Hello everyone! In today's video, we're excited to introduce an innovative solution that lowers the barrier to Physical AI research: OMY and Physical AI Tools. 👉 Introducing OMY Our OMY is a 6-axis manipulator & gripper specifically designed for Physical AI research. Notably, its Leader-Follower configuration provides an optimized environment for Imitation Learning. Equipped with collision detection and gravity compensation, OMY enables safe and efficient acquisition of training data. 🛠️ Introducing Physical AI Tools Physical AI Tools is a user-friendly software developed to allow anyone to easily participate in Physical AI research. Built upon ROS 2 and LeRobot, it offers intuitive UI-based data acquisition and monitoring without complex procedures. It's designed to be easily usable by beginners in robotics learning and part-time data acquisition specialists. Furthermore, with Web UI support, it can be accessed from various devices like mobile phones and tablets. 💡 What You'll See in This Video: In this video, we'll walk you through the entire workflow, from introducing OMY to the detailed processes of data acquisition, training, and inference using Physical AI Tools. For the inference, we are currently utilizing the ACT Policy model. ✨ The Future of Physical AI Tools: While currently focused on data acquisition and inference, we will soon be adding training functionalities. We also plan to integrate various open-source models like NVIDIA GR00T N1.5 to provide even more powerful features. All the code for our Physical AI Tools is open-source, emphasizing transparency and extensibility. We kindly ask for your continued interest and anticipation for our future contributions to the advancement of Physical AI research. Thank you so much for watching the video! Please feel free to leave any questions in the comments. 📥 GitHub Repo - https://lnkd.in/gafsrpqM - https://lnkd.in/gdnJg695 📺️ Video - https://lnkd.in/gwhaG7mT #ROBOTIS #OMY #ROS #NVIDIA #HuggingFace #LeRobot #robotics #physicalai #ai #artificialIntelligence #robot #opensource #robotarm #LeaderFollower #ImitationLearning

Are you interested in Physical AI? We've been conducting various tests and related work since last year. Among them, the newly released Physical AI Tools significantly improve upon the inconvenient CLI-based Physical AI operations by providing a user-friendly Web GUI. It's compatible with both our AI Worker and OMY products. With its intuitive UI, Physical AI Tools supports data acquisition and inference without complex procedures. Built on ROS 2 and Hugging Face LeRobot, it's accessible via Web UI from various devices and designed to be easily used by beginners. All content is open-source. We've currently tested it with ACT, PI0, OpenVLA, and GR00T N1.0, and we look forward to Physical AI Tools becoming even more powerful with future integrations of various open-source models like NVIDIA GR00T N1.5 and Google Gemini Robotics. 📥 GitHub Repositories https://lnkd.in/gafsrpqM https://lnkd.in/gdnJg695 📺️ Video https://lnkd.in/gwhaG7mT #ROBOTIS #OMY #ROS #NVIDIA #HuggingFace #LeRobot #robotics #physicalai #ai #artificialIntelligence #robot #opensource #robotarm #LeaderFollower #ImitationLearning

At KIMLAB (Kinetic Intelligent Machine LAB) at University of Illinois Urbana-Champaign, we have been developing PAPRAS (Plug-And-Play Robotic Arm System), a shareable and reconfigurable robotic arm platform. We explored a wide range of applications with different embodiments and had extensive discussions on designing a unified control interface, effective data representation, and efficient data collection strategies. Building on this foundation, we introduce PAPRLE (Plug-And-Play Robotic Limb Environment), a modular ecosystem that enables flexible placement and intuitive control of robotic limbs. PAPRLE allows users to dynamically reconfigure limb arrangements and control them through various input modalities, including puppeteering systems, gaming controllers, and VR-based interfaces. This flexibility supports a wide range of teleoperation scenarios and enables seamless adaptation to different task requirements. PAPRLE also supports bilateral teleoperation, independent of the type or configuration of the follower robot. By enabling both joint-space and task-space control with real-time force feedback, the system enhances user fidelity and physical interaction awareness, making it suitable for complex and responsive manipulation tasks. Obin Kwon and Sankalp Yamsani have been leading this project, with great support from Noboru Myers, Sean Taylor, Jooyoung Hong. We also had valuable collaboration with kyungseo park at DGIST (Daegu Gyeongbuk Institute of Science and Technology) and Alex Alspach at Toyota Research Institute. We are now releasing the PAPRLE project page alongside the PAPRAS project page, providing detailed information and access to the codebase. Please explore the project pages and try out various PAPRAS applications through the playground page. - PAPRLE (Plug-And-Play Robotic Limb Environment) Project page: https://lnkd.in/ghTMh_9Q Youtube : https://lnkd.in/g-VB-WqT - PAPRAS (Plug-And-Play Robotic Arm System) Project page: https://lnkd.in/gh4FD4_k Playground: https://lnkd.in/ga7E6QwH