Northrop Grumman is turning to NVIDIA to take the next leap in autonomous space systems.
The defense giant will use NVIDIA’s Omniverse platform and Isaac Lab to simulate and train AI-powered robotic systems for spacecraft that can think, adapt, and act on their own.
The aim is to create a new class of “cognitive” spacecraft capable of performing tasks like docking, servicing, and in-orbit repair with minimal human oversight.
At the core of this shift is agentic AI: systems that don’t just react, but proactively plan missions, adapt to changing conditions, and carry out operations across all phases of flight.
“By applying our extensive space knowledge and mission expertise to NVIDIA’s advanced AI technologies, we are poised to deliver innovative and high-assurance AI to enable new, more complex missions, with faster planning and significantly lower operating costs,” said Rob Fleming, corporate VP and president of Northrop Grumman Space Systems.
AI Tools
Tailored AI tools will enhance Northrop Grumman’s space projects by improving technology performance in dynamic environments, accelerating development, and boosting mission readiness.
NVIDIA Omniverse blends generative AI with 3D simulation, letting engineers build and test systems in realistic virtual environments before deployment.
Isaac Lab, part of NVIDIA’s Isaac robotics platform, is designed to train robotic systems through simulation.
It helps robotic systems “learn” how to operate effectively and adapt to new tasks with greater autonomy.
Agentic AI takes autonomy further, using advanced machine learning to understand goals and act independently to address them.
Unlike generative AI, it executes multi-step processes and handles complex tasks without human intervention.
In the long term, these technologies could power fully autonomous spacecraft and streamline current missions by automating routine tasks.
