Articles tagged with "dexterous-manipulation"
Video: 'Backdrivable' robot hand spins nut on bolt at incredible speed
Kyber Labs, a New York-based robotics company, has introduced a robotic hand capable of spinning a nut on a bolt at exceptionally high speeds in real time, without any video edits. This performance is enabled by the hand’s fully backdrivable and torque-transparent actuators, which allow it to naturally adapt to the nut’s movement. The backdrivability means external forces can move the robot’s fingers, and the system can infer torque from motor current, eliminating the need for complex tactile sensors. This design philosophy aims to simplify manipulation tasks, making control software and AI learning systems more reliable and efficient by offloading variability handling to the hardware itself. The robotic hand mimics human-like mechanical compliance and precision, enabling fluid and dexterous manipulation suitable for delicate tasks at scale. Kyber Labs emphasizes that general-purpose robotic hands remain a significant bottleneck in advancing robot capabilities, particularly for complex assembly and manufacturing operations. Their platform includes dual arms with human-like hands designed specifically for embodied AI, facilitating large
robotroboticsrobotic-handbackdrivable-actuatorAI-based-controlrobotic-manipulationdexterous-manipulationVideo: Robotic hand with 16 joints perfectly opens soda can like humans
TetherIA, a California-based company, has developed the Aero Hand Open, an open-source, tendon-driven robotic hand designed to tackle the challenge of dexterous manipulation in robotics. Featuring 16 joints, the robotic hand can perform precise tasks such as grasping small objects, picking up an iPhone, and even opening a soda can with human-like dexterity. Its underactuated design uses cables running through the fingers instead of individual motors for each joint, allowing the hand to conform naturally to various object shapes. This compliance enhances its ability to handle everyday objects with precision and adaptability. The Aero Hand Open emphasizes simplicity, affordability, and accessibility. It is fully 3D-printable, lightweight, and integrates open-source ESP32-based firmware and ROS2-compatible control software, making it suitable for research and educational use. Priced at $314 for a complete kit, it offers a modular platform that can be assembled with standard equipment and off-the-shelf components. Additionally, the system includes a
robotroboticsrobotic-handdexterous-manipulationopen-source-roboticstendon-driven-actuation3D-printingWatch: South Korean robotic hand performs a moonwalk with its fingers
South Korea’s Hanyang University showcased a playful yet technically impressive demonstration of robotic dexterity by having the Tesollo DG-5F robotic hand perform a finger-based moonwalk to Michael Jackson’s "Smooth Criminal." This 40-second video, created by PhD candidate Ji-Hun Bae at the RoCogMan Lab, serves as an academic parody inspired by a similar moonwalk performed by KAIST’s humanoid robot using legs. The demonstration highlights the robotic hand’s fine motor control and joint coordination, emphasizing that not only humanoid robots but also robotic fingers can exhibit dance-like movements. The Tesollo DG-5F is a fully actuated, five-finger anthropomorphic robotic hand with 20 degrees of freedom, designed for human-like precision. Each joint is independently controlled, enabling complex, coordinated movements and various grip patterns such as pinch, power, and precision grips. Equipped optionally with high-precision force sensors, the DG-5F provides tactile feedback essential for adaptive force control and slip
roboticsrobotic-handdexterous-manipulationanthropomorphic-robotTESOLLO-DG-5Frobotic-cognitionmedical-roboticsGoogle DeepMind introduces on-device Gemini AI model for robots - The Robot Report
Google DeepMind has introduced Gemini Robotics On-Device, a new AI model designed for two-armed robots that enables general-purpose dexterity and rapid task adaptation while operating locally with minimal computational resources. This vision language action (VLA) model brings the multimodal reasoning and real-world understanding capabilities of Gemini 2.0 into physical robotics, allowing robots to perform complex tasks such as unzipping bags or folding clothes. The on-device nature of the model ensures low-latency inference and robustness in environments with limited or no network connectivity, making it suitable for latency-sensitive applications. Alongside the model, DeepMind released a software development kit (SDK) to help developers test and fine-tune the model for specific tasks and environments, requiring as few as 50 to 100 demonstrations. DeepMind has demonstrated the adaptability of the Gemini Robotics On-Device model beyond its initial training on ALOHA robots by successfully applying it to other platforms, including the bi-arm Franka FR3 robot and the
roboticsAI-modelon-device-AIdexterous-manipulationGemini-RoboticsDeepMindrobot-SDKOctopus-inspired robot can decide how to grip objects with accuracy
robotoctopus-inspiredsoft-roboticssuction-intelligencedexterous-manipulationenvironmental-sensingmultimodal-perception