Google DeepMind has introduced two artificial intelligence (AI) models designed specifically to enhance robots' capabilities. The advancements in conversation, movement, and understanding are highly promising.
The convergence of robotics and AI continues to grow, with companies making significant strides in integrating AI into humanoid robots, revitalizing a sector that appeared stagnant. These new models are based on Gemini 2.0, Google's advanced language model, and aim to further this integration.
The first model, Gemini Robotics, is an advanced vision-language-action (VLA) model, enabling robots to perform physical actions and adapt to various situations. It aims to create "more efficient, responsive, and resilient robots" in changing environments. Meanwhile, Gemini Robotics-ER serves as a sophisticated interpreter between AI and the physical world, allowing machines to understand spatial relations like humans, thereby automating instruction generation after observing human demonstrations.
