Nvidia introduces GR00T to build Intelligent Humanoid Robots

Nvidia now sets its eye on humanoid robots, as they introduce a new GR00T to empower the AI to a whole new level. The chip manufacturer just announced Project GR000T or as it is known as Generalist Robot 00 Technology. What is Generalist Robot 00 Technology and how it is going to influence the world of robotics? 

Nvidia promises to start a new wave for humanoid robots by introducing GR00T. Humanoid robots are nothing new in the world, but none of the humanoids up until this point had a difficult time carrying out basic tasks. While the movie industry has pictured humanoids to be capable in almost everything, reality begs to differ.

But the Nvidia project aims to strengthen the humanoid robots with human-like understanding and enable AI to carry out proper movement patterns. If you are following the AI news in the world, you would know that Figure AI, another AI robotics company, showcased their humanoid robot powered by ChatGPT.

We saw Figure AI following instructions and helping with tasks such as cleaning up and handing over objects.

What is GR00T?

It is basically a foundation model for humanoid robots, it will empower the humanoid robots to be far too efficient in the real-work space environment. The inclusion of AI and other tools, more on that later, is set to make things move faster.

Project GR00T provides a platform for robots to help them understand natural language, recreate movements by witnessing human actions. The observation goes far beyond just reenacting the basic movements.

Nvidia will also allow new training methods that will allow the robots to also obtain new skill sets such as coordination, dexterity, and other handful of ways to get better in the work field.

Nvidia CEO, Jensen Huang showed multiple videos of robots accomplishing many of the common tasks of our everyday life. He went on this to say about the AI in robotics –

“Building foundation models for general humanoid robots is one of the most exciting problems to solve in AI [Artificial Intelligence] today,” He continued by adding, “The enabling technologies are coming together for leading roboticists around the world to take giant leaps towards artificial general robotics.”

Some of the footage of robots doing mundane activities here shares some similarities with the Kinect presentation from Microsoft.

GR00T Training

The new model here runs on Nvidia GPU-accelerated simulation, which enables the humanoid to pick up human imitation from simple observation. There is also the Nvidia Isaac Lab for reinforcement learning, which also further develops the basic understanding in robotics.

There are basically two types of training models followed to strengthen the humanoid’s understanding. One is known as Imitation learning and the other one is known as reinforcement learning.

GR00T Imitation Learning

Here the robot observes an expert handling a task, so it can carry out the actions by imitating what it saw.

GR00T Reinforcement learning

This kind of training relies on machine learning methods as it trains the software to make haste decisions to get to the optimal results.

The idea of GR00T is to hone the mind capability of AI-enabled systems, it allows the humanoids to learn new skills and reenact everyday actions properly just like humans.

GR00T is a fascinating piece of tech, as of now, many leading names in robotics such as Boston Dynamics, Sanctuary AI, Unitree, Agility Robotics, and many more have shaken the hands of Nvidia to further implement GR00T.

Sanctuary AI CEO and co-founder Geordie Rose had this to say about the including Nvidia to their vision – 

“Embodied AI will not only help address some of humanity’s biggest challenges, but also create innovations which are currently beyond our reach or imagination. Technology this important shouldn’t be built in silos, which is why we prioritize long-term partners like NVIDIA.”

The story doesn’t stop here. As now Nvidia also made Isaac better, now it comes with multi camera 3D vision. Nvidia also added Issac manipulator and isaac Perception, both of them carry out different operations for robots. 

While the manipulator helps the robots to manipulate the object around its surroundings, the job of Isaac perceptor deals with nothing but Perception in patterns.

Nvidia also released a new computer called Jetson Thor, it runs on Nvidia’s Thor system-on-chip or SOC. The basic work of Thor is to further empower the intelligence of the robots, the chip here is also made to handle convoluted tasks in mind. That’s not all, Thor will also help robotics to carry out the tasks while keeping the safety of humans in mind.

Thor will make sure the natural interaction between humans and robots stays in the safety zone. Nvidia is showing no sign of stopping when it comes to AI, as recently they added AI to further the technology behind 6G networking. Now with robotics, AI is here to stay whether we like it or not.

What does the humanoid robot mean for the world?

The question of AI taking over jobs, is not some distant timeline in the future, it is imminent. Honing the robots to perform common and complex tasks in the workspace can only lead to replacement of many white collar professions.

Humanoid robots will easily carry out the repetitive tasks without losing any precision in the process. This is something one can expect to see in a script for a sci-fi movie, but now it is being played out in real-life. The evolution of AI has further amplified the need to go full on robotics for many industries.

For more on Tech, Follow Tecxology