Nvidia and Google DeepMind will help power Disney’s cute robots

Nvidia, Disney, and Google DeepMind collaborate on Newton, a robotics physics engine.

: Nvidia is collaborating with Disney Research and Google DeepMind to develop Newton, a physics engine designed to simulate robotic movements in reality. Disney will utilize Newton to power advanced entertainment robots, like BDX droids inspired by Star Wars showcased at GTC 2025. Jensen Huang introduced Groot N1, a new AI model for humanoid robots, and detailed future plans for next-gen AI chips and personal AI computers. Newton aligns with Google DeepMind's MuJoCo engine, enhancing robot expressiveness and task handling precision.

Nvidia, a leading technology company specializing in graphics processing units and AI, announced a collaboration with Disney Research and Google DeepMind at GTC 2025 to develop a new physics engine called Newton. Announced by Nvidia’s CEO, Jensen Huang, Newton aims to simulate robotic movements in real-world environments. Disney is closely involved with the initiative and plans to integrate Newton into its next-generation amusement park robots, such as the Star Wars-themed BDX droids. Jensen Huang highlighted this work by showcasing a BDX droid, which even made an appearance on stage at the conference.

Disney has been nurturing the concept of Star Wars-inspired robots for several years and has showcased them at various events, such as SXSW 2025. With the integration of the Newton engine, Disney is convinced that the technology is now sufficiently advanced to deploy these robots in its theme parks around the globe. Kyle Laughlin, Senior Vice President of Disney Imagineering, underscored the importance of Newton, describing it as a key component in driving future Disney entertainment robot technologies.

Newton is engineered to enhance the expressiveness and operational precision of robots. Nvidia posits that this physics engine will empower developers by simulating complex interactions that robots must navigate in the natural world, which often poses significant challenges. Developers can customize Newton extensively, allowing robotic programming for interactions with various objects such as food, fabric, and sand. This customizability places Newton among the most versatile tools for engineering realistic robotic environments.

The physics engine is designed to operate seamlessly within Google DeepMind’s broader suite of robotic development utilities, including its MuJoCo engine, which focuses on simulating robots with multiple joints. Nvidia’s introduction of Newton was among several strategic releases during GTC 2025. The company also launched Groot N1, a foundation model purposed for humanoid robots to enhance environmental perception and reasoning. Additionally, Nvidia shared timelines for the introduction of their upcoming AI chips, namely Blackwell Ultra and Rubin, and introduced a line of personal AI computers.

The project with Nvidia and Google DeepMind aligns with Disney’s long-term strategic goals to incorporate cutting-edge technology in its entertainment offerings. It exemplifies a growing trend where AI and robotics converge to enhance customer experiences in entertainment. By utilizing advanced simulation tools such as Newton, companies like Disney can push forward in developing more lifelike and responsive robotic characters, adding a new dimension to their thematic storytelling abilities and park attractions.

Sources: Nvidia, TechCrunch, Disney

technologyAIrobotics