Google unveils DeepMind's advanced Gemini Robotics 2 model series for humanoid robots
deepmind gemini google robotics
| Source: SiliconANGLE | Original article
Google DeepMind introduces Gemini Robotics 2 model series for humanoid robots. The new models are designed to power advanced robotics capabilities.
Google DeepMind has debuted the Gemini Robotics 2 model series, a family of models designed to power humanoid robots. This new series enables whole-body intelligence, advanced dexterity, and coordination among multiple robots in shared spaces. The Gemini Robotics 2 models can control a full humanoid robot, from its feet to its fingertips, and also support bi-arm robots with enhanced manipulation capabilities.
This development matters because it addresses a significant limitation in learning-based robotics: the separation of locomotion and manipulation. Typically, robots rely on separate models for navigation and manipulation, but Gemini Robotics 2 unifies these functions under a single vision-language-action model. This unified approach has the potential to improve the overall performance and versatility of humanoid robots.
As the field of robotics continues to evolve, it will be interesting to watch how the Gemini Robotics 2 series is applied in various contexts, such as research, industry, and healthcare. The ability to control and coordinate multiple robots in shared spaces could lead to significant advancements in areas like manufacturing, logistics, and assistive care. With this debut, Google DeepMind is taking a significant step towards general, useful robotics, and the outcomes of this technology will be worth following in the coming months.
Sources
Back to AIPULSEN