Google Integrates DeepMind AI Into Robotics

On Mar 12, 2025

Google is integrating its DeepMind artificial intelligence (AI) technology into robotics, bringing advancements in AI to the physical world.

Introducing two new AI models from @GoogleDeepMind, based on Gemini 2.0, that lay the foundation for a new generation of helpful robots: Gemini Robotics, a vision-language-action model adds physical actions as a new modality, and Gemini Robotics-ER, a vision-language model that… https://t.co/i7ubkIR6s1

— Google (@Google) March 12, 2025

On Wednesday, the company unveiled two new AI models—Gemini Robotics and Gemini Robotics-ER (Extended Reasoning)—both powered by Gemini 2.0, which Google describes as its “most capable” AI to date.

Unlike previous AI models focused on generating text and images, Gemini Robotics enables physical action commands for robot operation.

The technology marks a significant step forward in the use of generative AI beyond digital outputs.

In a blog post, Google announced its partnership with Texas-based robotics developer Apptronik to build the next generation of humanoid robots using Gemini 2.0.

Apptronik, known for collaborations with Nvidia and NASA, welcomed Google’s involvement, which comes after the company joined Apptronik’s $350 million funding round last month.

Demonstration videos showed Apptronik robots performing tasks such as plugging a device into a power strip, packing a lunchbox, moving plastic vegetables, and zipping up a bag, all in response to spoken commands. Google, however, has not provided a timeline for commercial availability.

Google outlined three essential qualities for AI-powered robotics: adaptability to various situations, responsiveness to instructions and environmental changes, and dexterity for precise physical tasks such as object manipulation.

Gemini Robotics-ER is tailored for roboticists, offering a foundation for training custom models. In addition to Apptronik, trusted testers such as Agile Robots, Boston Dynamics, Agility Robots, and Enchanted Tools will have access to this model.

Other tech giants are also pursuing advances in AI-powered robotics. OpenAI recently invested in Physical Intelligence, a robotics startup, and hired Meta’s former augmented reality leader to spearhead its hardware initiatives. Similarly, Tesla is advancing its Optimus humanoid robot project.

Google CEO Sundar Pichai emphasized the importance of robotics in applying AI advancements to real-world scenarios.

Pichai noted that Google’s multimodal AI models would enable robots to adapt dynamically and respond to their surroundings effectively.

The initiative signals a transformative shift in the integration of AI technologies into physical systems, promising significant advancements in robotics and real-world applications.

CNBC