Google has introduced Google Gemini Robotics, two new artificial intelligence (AI) models, to enhance robot capabilities. Built on the Gemini 2.0 foundation, these models improve how robots understand instructions and complete complex tasks.
Enhancing Robot Intelligence
In a blog post on Wednesday, Google announced that Google Gemini Robotics enhances vision-language-action (VLA) abilities. This allows robots to handle new environments and objects without prior training. The model integrates AI to make robots more adaptable and efficient.
Gemini Robotics-ER: Spatial Awareness and Planning
Google also introduced Google Gemini Robotics-ER, which focuses on spatial reasoning and task planning. This model helps robots understand surroundings, estimate states, and generate real-time execution code. According to Google, the innovation allows roboticists to connect AI with existing control systems.
Compatibility with Robotic Platforms
The models were trained using the ALOHA 2 bi-arm robotic platform. However, they can also operate on Franka arms and humanoid robots like Apollo, developed by Apptronik. Google has partnered with Apptronik to push advancements in humanoid robotics.
Collaborations and Industry Impact
Google is also working with major robotics firms, including Agile Robots, Agility Robotics, Boston Dynamics, and Enchanted Tools, to test Gemini Robotics-ER. These collaborations aim to improve AI-powered automation across industries. The models enhance efficiency, adaptability, and human-robot collaboration.
Future of AI in Robotics
With this, the future of AI-driven automation looks promising. These models set the stage for more intelligent, capable, and versatile robots in various industries.
