Recent advancements in robotics have led to significant breakthroughs, including the development of a 4-legged robot capable of performing parkour and a miniature wheeled robot designed to navigate through tiny, complexly shaped pipes by converting fluid flow into mechanical energy. The quadrupedal robot ANYmal, developed by researchers at ETH, has been taught new skills through machine learning, enabling it to overcome obstacles and navigate through challenging scenarios without prior knowledge of the environment. This training involved programming modeled after human experts in parkour and the use of the Dreamer algorithm, allowing ANYmal to learn how to walk directly in the physical world in only 1 hour. Additionally, soft robots have been engineered to grow and shapeshift, achieving complex motions through a system of fluid-driven actuators. These innovations highlight the diverse applications and capabilities of modern robotics in overcoming physical challenges and adapting to various environments.
Agile Omnidirectional Mobile Robot with Gravity-Compensated Wheel-Leg Mechanisms for Human Environment Video Credit: IRIM LAB KOREATECH #robotics #engineering #technology #robots #roboticengineering #research https://t.co/YTPMib2Oab
Our soft robot arm is built with a unique design, making it an ideal research tool for vision-based manipulation and manipulation in unstructured environments due to its adaptability. #AI #Manipulation #vision #LLM #robotarm #MachineLearning https://t.co/V1IoxGofnN
Soft robots harness viscous fluids for complex motions A team of researchers designed a new—and surprisingly simple—system of fluid-driven actuators that enable soft robots to achieve more complex motions. Video source: Science X https://t.co/ubyXp3pJ8Z
Engineers develop soft robot that can grow and shapeshift https://t.co/Ftyeo0dFrV
#ANYmal is usually focused on its task of inspecting industrial facilities, but this time it had a training day. Thx to the @leggedrobotics team, ANYmal conquers challenging scenarios. This allows for path planning without prior knowledge of the environment. https://t.co/1FCMZhuz5e
‘Parkour’ robot dog can leap, jump, and crawl its way through complex obstacle courses https://t.co/BI0TvkfU4a
Researchers trained a quadruped robot to learn how to walk directly in the physical world without simulators. Learning from scratch in only 1 hour was possible by using the Dreamer algorithm to continuously learn a world model and plan inside of it. https://t.co/mDxqJvrGT1
ANYmal, a dog-inspired robot, learned to do parkour thanks to programming modeled after human experts. Learn more in our daily newsletter, #ScienceAdviser: https://t.co/DMRQk6cZLg https://t.co/b2uV17jwv2
ANYmal, a dog-inspired robot, learned to do parkour thanks to programming modeled after human experts. Learn more in @ScienceMagazine’s daily newsletter, #ScienceAdviser: https://t.co/0IYopJQFbC https://t.co/D2seLUZDjc
The quadrupedal robot ANYmal went back to school and has learned a lot. @ETH researchers used machine learning to teach it new skills: the robot can now climb over obstacles and successfully negotiate pitfalls. https://t.co/HRTs9MyGGh #robotics #tech #engineering #robots #AI https://t.co/vSDY6PHYfM
A miniature wheeled robot can convert fluid flow into mechanical energy, enabling it to move through tiny, complexly shaped pipes with varying diameters. Learn more in @SciRobotics: https://t.co/C2DDZizL8Z https://t.co/ExQiIfdrNv
This 4-Legged Robot From Switzerland Does Parkour https://t.co/Jm2Ty0HhIh #Robotics #Robots #RobotDesign