California – Google has introduced Gemini Robotics and Gemini Robotics-ER, two new AI models designed to improve how robots think, move, and interact with the world. They wants to create smarter, more adaptable robots that can perform real-world tasks with better accuracy and understanding.
Advancing AI with Embodied Reasoning
Google thinks, the Artificial Intelligence could be very useful, it must not only process information but also understand and respond to its surroundings, just like humans can do. This concept, known as “embodied reasoning,” which allows robots to make decisions based on real-world observations rather than programmed/coded for specific tasks
Gemini Robotics is built on platform of Google’s Gemini 2.0 AI model, Gemini Robotics introduces physical actions as an output. Because of this model, robots are now able to act on what they see and hear. This advancement brings AI one step closer to functioning like a human in everyday evironments.
Gemini Robotics: Smarter Robots for Everyday Use
Gemini Robotics combines vision, language, and movement, making it more interactive and precise. It can process real-time instructions, adapt to changes, and complete complex tasks, such as picking up objects, folding paper, or opening a bottle.
One of its key strengths is its flexibility, while it was primarily trained on the bi-arm robotic platform, ALOHA 2, it also works with Franka arms, which are commonly used in research labs. This adaptability makes it suitable for different types of robots and applications, from homes to industrial settings.
Introducing Gemini Robotics-ER for Better Spatial Awareness
Alongside Gemini Robotics, Google has launched Gemini Robotics-ER, which enhances how robots understand space, objects, and motion. This model improves 3D object recognition and movement planning, helping robots pick up objects more accurately and move safely in their environment.
For example, if a robot is shown a coffee mug, Gemini Robotics-ER can determine the best way to grasp it and plan the safest path to pick it up. This feature makes AI-powered robots more reliable for real-world tasks, reducing errors and improving efficiency.
The Future of AI-Powered Robotics
Google’s advancements in robotics mark a significant step toward creating robots that can work alongside humans in various industries, including healthcare, manufacturing, and home assistance. The company also emphasizes safety, ensuring that these robots operate responsibly and avoid risks.
With continued improvements, Gemini Robotics and Gemini Robotics-ER could lead to a future where AI-powered robots assist in everyday tasks, making life easier and work more efficient. Google is trying to work with researchers and companies to refine these models and bring them closer to real-world applications.
Megan Davies is a reporter for White Pine Tribune. After graduating from the University of Toronto, Megan got an internship at the CBC News and worked as a reporter and editor. Megan has also worked as a reporter for Global Toronto. Megan covers economy and community events for White Pine Tribune.
