Tesla has further developed its humanoid robot Optimus. The robot completes in a video clip on X, formerly Twitter, a few yoga poses to demonstrate the capabilities of the actuators and the robot’s “sense of balance.” In addition, he is now able to sort terminal blocks by color independently.
Advertisement
The movement routines begin with self-calibration of the arms and legs. This works through image processing systems that recognize and determine the position of the limbs in space. They also enable the Optimus robot to learn new skills.
The Tesla Optimus now moves much more smoothly than it did just a few months ago. He can balance his weight so that he can do yoga exercises on one leg.
Recognize and sort building blocks
The image processing systems on board allow the robot to learn new tasks. This is a prerequisite for being able to be used as general-purpose robots in factories. A neural network enables Optimus to detect different colored terminal blocks and sort them by color. It can also react to changes that a person makes during the sorting process, for example by swapping building blocks. The robot recognizes the changes made and picks out the appropriate building block and places it correctly in the correct position. The actuators on the robot’s hands are very flexible. The hands appear about as dexterous as human hands.
Tesla introduced its Optimus robot in October 2022. Back then, the robot staggered quite clumsily across the stage at the Tesla AI Day. Tesla has now made various improvements that enable the robot to move faster and more fluidly – last seen in May 2023. Previously, Tesla also showed the first approaches to how the robot can interact with objects.
Tesla boss Elon Musk sees great potential for the Optimus. He believes he could shift Tesla’s core business from electric cars to general-purpose robots. Musk predicted that demand could be 10 to 20 billion units. The robot should then cost less than $20,000.
(olb)
To home page
#Humanoid #robot #Tesla #Optimus #yoga #sort #building #blocks #autonomously