NVIDIA Is Working With More Than Just Agile Robots: Its Growing Robotics Partnerships Explained
Something important is happening in the global robotics industry.
Robots are no longer limited to performing the same repetitive movement on a factory production line. Companies are developing machines that can walk, pick up objects, understand their surroundings, and perform tasks in warehouses, factories, hospitals, and other environments.
But building an intelligent robot requires more than a mechanical body.
A robot also needs computing hardware, artificial intelligence, software, simulation, sensors, and systems that help it understand the physical world.
This is where NVIDIA comes in.
On March 16, 2026, NVIDIA announced a broad collaboration with companies across the robotics industry. Its ecosystem includes humanoid robot developers, industrial robotics manufacturers, AI software companies, and other technology businesses.
The companies named in NVIDIA's announcement include Figure, Boston Dynamics, Agility, Agile Robots, NEURA Robotics, ABB Robotics, FANUC, and KUKA.
NVIDIA is not working exclusively with Agile Robots. It is building a much larger robotics ecosystem in which different companies use NVIDIA computing and software technologies to develop, train, simulate, and deploy intelligent machines.
Here is what each major collaboration means, where the companies come from, and why these partnerships matter.
What Is NVIDIA Trying to Achieve With Robotics?
NVIDIA is best known for its GPUs, which are widely used to train and run artificial intelligence systems.
However, the company is expanding its role beyond computers and data centers.
Its robotics strategy focuses on what NVIDIA calls Physical AI — artificial intelligence that can understand and interact with the physical world.
For example, a chatbot can answer a question using text. A physical AI system must do something more complicated: recognize an object, understand where it is located, plan a movement, control a robot's joints, and respond if something changes.
NVIDIA provides technologies that can support different stages of this process:
- NVIDIA GPUs and Jetson computing: Provide computing power for AI training and real-time inference.
- NVIDIA Isaac: A collection of robotics development and simulation technologies used to develop, train, and test robots.
- NVIDIA Omniverse: Helps companies create physically accurate virtual environments and digital twins for simulation.
- NVIDIA Cosmos: Provides world-model technology intended to help AI systems learn and reason about physical environments.
- NVIDIA Isaac GR00T: Provides AI models and development resources for humanoid robotics.
The purpose is to help robotics companies develop intelligent machines more efficiently, test them in virtual environments, and move their AI systems toward real-world deployment.
NVIDIA does not necessarily manufacture the complete robot in these collaborations. Many of its partners build the robot itself, while NVIDIA contributes computing hardware, software, AI models, and simulation tools.
1. Figure — United States
Country: United States
What it does: Figure develops humanoid robots designed to perform tasks in environments built for people.
Figure is developing robots that can move through workplaces, handle objects, and potentially perform a range of physical tasks.
Why is Figure working with NVIDIA?
Humanoid robots need to coordinate several complex abilities at the same time.
They must maintain balance, recognize objects, understand their surroundings, and control their movements. Developing these capabilities through real-world testing alone can be expensive and slow.
NVIDIA's robotics technologies can help Figure:
- Train robot AI using simulation.
- Test movements and manipulation in virtual environments.
- Develop perception and reasoning capabilities.
- Accelerate the development of humanoid robot systems.
The broader goal is to help Figure develop robots that can perform useful work in real-world environments.
The collaboration does not mean NVIDIA is manufacturing Figure's humanoid robots. Figure remains the robot developer, while NVIDIA supplies technologies that can support the development and training of intelligent machines.
2. Boston Dynamics — United States
Country: United States
What it does: Boston Dynamics develops advanced robots, including Spot, Atlas, and Stretch.
Its robots are known for their mobility, balance, and ability to operate in environments that can be difficult for traditional machines.
Why is Boston Dynamics working with NVIDIA?
Boston Dynamics has already developed sophisticated robot hardware and control systems. The next challenge is making robots more intelligent and capable of handling a wider variety of tasks.
NVIDIA's simulation and AI technologies can help Boston Dynamics develop and validate robot behaviors before testing them in the real world.
For example, developers can use virtual environments to investigate how a robot should move around obstacles, interact with objects, or respond to different situations.
The collaboration supports the development of more capable physical AI systems, including humanoid and other advanced robots.
It is important to understand that Boston Dynamics' robots are not all humanoids. The company also develops quadruped robots and warehouse automation systems.
3. Agility Robotics — United States
Country: United States
What it does: Agility Robotics develops Digit, a humanoid robot designed for work in logistics and industrial environments.
Digit has a human-like form with legs and arms, allowing it to move through spaces designed for people and handle certain physical tasks.
Why is Agility Robotics working with NVIDIA?
Agility Robotics needs its robots to understand their surroundings, move safely, and perform tasks reliably.
NVIDIA's AI models, simulation tools, and computing infrastructure can help Agility Robotics:
- Train robots in virtual environments.
- Improve movement and object-handling capabilities.
- Test robot behavior before physical deployment.
- Develop AI systems that can adapt to changing environments.
The purpose is to accelerate the development of humanoid robots that can work in logistics, manufacturing, and other industrial settings.
This collaboration is part of NVIDIA's broader effort to support multiple humanoid robot manufacturers rather than relying on one robot company.
4. Agile Robots — Germany
Country: Germany
What it does: Agile Robots develops intelligent industrial robots and automation systems. Its robotics portfolio includes systems designed for flexible manufacturing, assembly, and handling tasks.
Agile Robots is different from Agility Robotics. Despite their similar names, they are separate companies from different countries.
Why is Agile Robots working with NVIDIA?
Industrial environments often contain tasks that are difficult to automate using traditional fixed programming.
A robot may need to recognize different objects, adjust its movements, or respond to changing production conditions.
NVIDIA's technologies can help Agile Robots develop AI-powered systems through simulation, training, and real-time computing.
The goal is to make industrial robots more adaptable, allowing manufacturers to automate a wider range of tasks without having to redesign the entire production process for every change.
Agile Robots also participates in NVIDIA's broader robotics safety ecosystem. NVIDIA announced work with partners on technologies intended to support safer deployment of physical AI systems.
5. NEURA Robotics — Germany
Country: Germany
What it does: NEURA Robotics develops cognitive robotics systems, including humanoid and collaborative robots.
Its robots are designed to interact with people and work in industrial environments. The company focuses on combining robotic hardware with sensing, perception, and intelligent control.
Why is NEURA Robotics working with NVIDIA?
NEURA Robotics needs AI systems that can understand objects and environments while controlling physical movements.
NVIDIA's Isaac and Cosmos technologies can help developers simulate and train robots, reducing the need to conduct every experiment using physical machines.
NEURA Robotics is also among the companies adopting NVIDIA's GR00T family of robotics models to accelerate the development of industrial humanoid applications.
The purpose is to develop robots that can perform more complex tasks, handle objects more effectively, and operate in real industrial settings.
6. ABB Robotics — Switzerland and Sweden
Company origin: ABB was formed through a merger of Swedish and Swiss engineering companies in 1988. Its headquarters are in Switzerland, while ABB Robotics has major historical roots in Sweden.
What it does: ABB Robotics develops industrial robot arms, automation systems, and software used in factories and manufacturing facilities.
Its robots are used in applications such as assembly, welding, material handling, and electronics manufacturing.
Why is ABB Robotics working with NVIDIA?
ABB Robotics and NVIDIA are working to improve how industrial robots are programmed, simulated, and deployed.
ABB is integrating NVIDIA Omniverse libraries into its RobotStudio simulation and programming environment.
This allows manufacturers to create virtual versions of robotic work cells and production systems, helping them test automation workflows before installing equipment in a real factory.
The collaboration can help manufacturers:
- Simulate robotic movements and factory layouts.
- Test production processes virtually.
- Identify potential problems before physical installation.
- Reduce the time and cost involved in commissioning automation systems.
The purpose is not simply to put AI inside a robot arm. It is to help manufacturers design and deploy entire robotic production systems more efficiently.
7. FANUC — Japan
Country: Japan
What it does: FANUC is a major industrial robotics and factory automation company. It manufactures industrial robot arms, CNC systems, and automation equipment used in manufacturing.
FANUC robots are widely used in applications such as automotive production, electronics manufacturing, and material handling.
Why is FANUC working with NVIDIA?
FANUC and NVIDIA are collaborating to bring more AI capabilities into industrial robotics.
FANUC is leveraging NVIDIA computing and simulation technologies, including Jetson edge modules, Isaac Sim, and Omniverse libraries.
These technologies can support:
- AI-powered robot perception.
- Virtual testing of manufacturing workflows.
- Simulation of robotic applications before deployment.
- Real-time AI processing within industrial environments.
The goal is to make industrial robots more intelligent and adaptable while helping manufacturers validate their automation systems before using them on production lines.
FANUC is an important example of how NVIDIA's robotics ecosystem extends beyond humanoid robots into traditional industrial automation.
8. KUKA — Germany
Country of origin: Germany
What it does: KUKA develops industrial robots and automation systems used in manufacturing, logistics, and other industrial applications.
Its robot portfolio includes machines for welding, assembly, material handling, and automated production.
Why is KUKA working with NVIDIA?
KUKA is integrating NVIDIA technologies into its industrial automation and simulation workflows.
NVIDIA's Omniverse libraries and Isaac simulation frameworks can help KUKA and its customers create digital twins of production environments.
This means companies can test robot movements, production layouts, and automation processes inside a virtual factory before deploying them physically.
The collaboration also supports AI-powered robotics by providing tools for simulation, training, and validation.
The objective is to help manufacturers deploy automation more quickly, improve production planning, and reduce the risk of costly changes after a factory system is installed.
KUKA originated in Germany, although its ownership has changed over time. It is currently part of China's Midea Group.
NVIDIA Is Also Working With Other Robotics Companies
The companies above are only part of NVIDIA's robotics ecosystem.
Its March 2026 announcement also included companies such as:
| Company | Country of origin | Robotics focus |
|---|---|---|
| AGIBOT | China | Humanoid and intelligent robots |
| Universal Robots | Denmark | Collaborative industrial robots |
| YASKAWA | Japan | Industrial robotics and automation |
| Hexagon Robotics | Sweden | Robotics and autonomous systems |
| Skild AI | United States | General-purpose robot AI |
| World Labs | United States | AI and spatial intelligence |
These companies have different roles. Some develop complete robots, some provide industrial automation systems, and others focus on AI software and models that can help robots understand and interact with the physical world.
The purpose of NVIDIA's ecosystem is to make its computing and software technologies useful across different types of robotics businesses.
Is NVIDIA Building Its Own Robots?
NVIDIA's robotics strategy is primarily focused on providing the technology that other companies use to build intelligent machines.
It develops computing hardware, AI models, simulation frameworks, and software tools that robotics manufacturers can integrate into their systems.
For example, Figure can use NVIDIA technologies to help develop a humanoid robot, while ABB can use NVIDIA Omniverse libraries to simulate industrial robot applications.
This approach allows NVIDIA to participate in several areas of robotics without needing to manufacture every robot itself.
In simple terms:
- Robot manufacturers build the physical machines.
- AI developers create the software that helps machines perceive, reason, and act.
- NVIDIA provides computing hardware and software technologies that can support the development, training, simulation, and deployment of physical AI.
Some companies perform more than one of these roles, and the exact technology used differs by collaboration.
What Happens Next?
NVIDIA's robotics ecosystem is likely to become increasingly important as more companies develop AI-powered machines.
Humanoid robots may be used for logistics and manufacturing tasks. Industrial robots may become more adaptable to changing production requirements. Simulation technologies may allow companies to test automation systems more extensively before installing them in factories.
However, these developments do not mean that every announced collaboration has already resulted in robots being deployed at scale.
The next important stage is proving that AI-powered robots can operate reliably, safely, and economically in real-world environments.
NVIDIA's strategy is to provide a common technology foundation that can help different robotics companies work toward that goal.
Conclusion
NVIDIA is not working only with Agile Robots. Its March 2026 robotics announcement highlighted a much broader ecosystem of companies, including Figure, Boston Dynamics, Agility Robotics, Agile Robots, NEURA Robotics, ABB Robotics, FANUC, and KUKA.
These companies come from different countries and specialize in different areas of robotics, from humanoid machines to industrial robot arms and factory automation.
NVIDIA's role is to provide AI computing, simulation, models, and software that can help these companies develop and deploy intelligent robots.
The broader story is that the robotics industry is moving toward a shared AI technology ecosystem, where robot manufacturers and computing companies work together to bring physical AI into factories, warehouses, and other real-world environments.
