A four-legged robot HUNTER managed to create a world record Guinness for the fastest 100 meter race completed by a machine of this type.
The brand, which is also considered a new milestone in robotic speedbecause the machine managed to travel this distance in time 19.87 seconds.
This goal was achieved after mastering the running technique reinforcement learningan artificial intelligence learning data mechanism with which the Hound’s system was programmed to maximize its engine capacity and increase speed.
At least this is confirmed by its developers, which they are Laboratory of design and control of dynamic robots of South Korea Advanced Institute of Science and Technology (KAIST)where HOUND achieved an average speed of 18.12 km/h (11.26 mph), breaking the previous limitations in four-legged locomotion robotics.
In addition, the robot weighs approx 45 kilos (£99.21) and was designed with special mechanics high-speed roboticswith its performance attributed to an advanced control system.
Designer Young Ha Shin points out that “all these movements were achieved with a single controller trained in the simulation through reinforcement learning”.
To achieve the most realistic experience possible, dynamic characteristics of the engine in simulations, suggesting potential for HOUND in future testing in real-world environments.
In addition, this machine is designed to adapt to different physical challengesit can run not only on a smooth surface, but also on terrain such as grass, overcome slopes of up to 22° and obstacles up to 35 cm high.
And it also features technology that allows for a wide range of motion facilitated by a belt-pulley system which optimizes the transmission of forces between the joints.
The choice of motors is critical, as expressed by Shin, who cautions that “instant and constant power are important characteristics when choosing a motor for dynamic leg locomotion.”
This distributes its force evenly and achieves a symmetrical and balanced movement of the four legs, which have been designed from lightweight materials to reduce their overall weight.
HOUND also integrates machine learning technologies and innovative mechanical design, and according to Professor Park Hae-won, who oversaw the development of the project, this breakthrough “proved that Korea has undisputed global technologies in robotic hardware and robotic self-driving artificial intelligence.”
tesla announced the arrival of a new version of its humanoid robot, the Optimus Gen 2designed with improved mobility options, especially in the neck and hands, as well as built-in sensors in your fingers that allow you to manipulate delicate objects using different levels of pressure.
Company led Elon Musk He further insists that Optimus Gen 2 can walk at a speed increased by 30% compared to its predecessor.
This advancement is largely due to a new configuration in the structure of the humanoid legs, which now have a shape more similar to human feet.
tesla The robot has also equipped force and joint sensors to achieve smoother and more natural movements.
In fact, their hands now not only have more fluidity in the movement of each finger, but are also equipped with sensors specialized to detect pressure.
What gives Optimus Gen 2 ability to carefully handle fragile objects such as egg either glass.