NVIDIA COMPUTEX 24 Related to Robots

NVIDIA COMPUTEX 24 Related to Robots

1. Omniverse for Physical AI and Robot Study to Change Robots' Future
  
- In the era of robotics, physical AI has a model for understanding and executing robot behavior, and LLMs are a major breakthrough in robot learning, perception, and action planning.
   - Robots using physical AI learn to perform all tasks automatically, and decision learning is performed by following the laws of physics in the virtual world.
   - In NVIDIA Omniverse, robots learn how to precisely manipulate objects or explore the environment autonomously, and minimize simulations to move the learned behavior as much as possible.
   - AI supercomputers, Jetson Orin and Jetson Thor, and Omniverse are needed to develop robots with physical AI, and the robots improve their technology in a simulated world.
   - It claims the importance of reinforcement learning, a fundamental technology in the field of robotics, and the future of innovating various industries using AI, and makes Omniverse the center of robot learning.

2. Complex ecosystems of robot warehouses and factories

   - The ecosystem of robotic warehouses is very complex and requires a number of companies, tools, and technologies to build modern warehouses. Warehouses are becoming more and more robotic, and it is expected that everything will be fully robotic soon.
   - In each of these ecosystems, SDK and API are connected to the software industry, SDK and API are connected to the edge AI industry and companies, and of course, there are systems for PLC and robot systems for ODM.
   - This is eventually integrated and created by the integrator, which builds warehouses for customers. Here is an example of KenMac building a robot warehouse for a large group.
   -The factory has a completely different ecosystem, Foxconn is building some of the most luxurious factories in the world. Their ecosystem includes edge computers and robots, software for factory design, robot programming, and PLC computers.

3. Robot factory innovation with adoption of NVIDIA-accelerated computing on Foxconn
   - Foxconn has used NVIDIA Omniverse and AI to build a robot factory against demand.
   - Factory planners use Omniverse to integrate facility and equipment data, especially with industry-led applications such as Siemens TeamCenter X and Autodesk REIT, and optimize future operations monitoring with Vision AI powered by NVIDIA Metropolis.
   - Using Omniverse's digital twin, Foxconn trains and tests robot development and NVIDIA Isaac AI and Metropolis AI applications.
   - Foxconn simulates two robot AIs on Omniverse to run NVIDIA Isaac and then runs them on the Jetson computer. Through this, Foxconn creates a robot factory that builds an NVIDIA AI supercomputer, through which it trains Foxconn's robot.

4. Innovation in the manufacturing industry through robot technology
   - MarkBest and BYD Electronics have leveraged *Isaac* technology to improve robot object recognition and human motion tracking.
   - Steelworks and Intrinsic introduced Isaac Perceptor and Manipulator for manufacturing innovation, respectively.
   - In addition, Gideon, Cargo Robotics, Salomon, and Techman Robot are successfully introducing *Isaac* technology to develop robot technology.
   - The robotics field is growing and is recognized as a promising field, which is why robotics technology is already being used in various fields such as *factory automation* and the production of human robots is expected to increase in the future.