Google’s “Dual-Model” Super Robot Brain Arrives, WiMi Invests in On-Device AI + Robotic Clusters

Google (GOOG) has announced the release of its latest Gemini Robotics 1.5 series of models, enabling robots to truly “think” and learn skills across diverse embodied forms.

The Latest Robotic “Brain” Model Arrives
DeepMind has released the Gemini Robotics 1.5 family of models for robotics and embodied intelligence, a new generation of “brains” specifically designed for robotics and embodied intelligence. This means that future robots will become intelligent partners that collaborate with humans and proactively complete complex tasks.
The Gemini Robotics 1.5 series includes Gemini Robotics 1.5 and Gemini Robotics-ER 1.5. Gemini Robotics 1.5, a state-of-the-art vision-language-action model, converts visual information and commands into motion commands for the robot to perform tasks.

Gemini Robotics-ER 1.5, a powerful vision-language model, can reason about the physical world, directly invoke digital tools, and create detailed multi-step plans to complete tasks. Combined, they form a powerful intelligent agent framework.

Google claims this breakthrough demonstrates the potential of “cross-embodied learning,” accelerating the learning of new behaviors and helping robots become more intelligent and practical. Robots in different scenarios (such as logistics and retail) can learn from each other, significantly accelerating the development of general-purpose robots.

The robotics industry is flourishing in multiple areas.

Undoubtedly, with the continuous integration and innovation of cutting-edge technologies such as artificial intelligence, embodied intelligence, and the Internet of Things, robots are transforming the traditional manufacturing industry at an unprecedented pace. As the name suggests, robots are still “machines.” Since the first industrial revolution, machines have been deeply embedded in daily life, becoming an inseparable necessity.

The Deep Integration of AI and Robotics: At the 2025 Automatica exhibition in Munich, Germany, the deep integration of AI and robotics hardware became a prominent feature. For example, Marc Segura, President of ABB’s Robotics division, stated that the next generation of multifunctional robots will feature “precise movement, semantic understanding, and 99% AI picking accuracy,” and will be able to flexibly switch between wheeled and legged configurations depending on the scenario. Kuka has deeply integrated generative AI with its SIM programming tools, enabling natural language programming and AI-vision collaborative grasping.

Humanoid robot technology is diverse: The United States focuses on functional learning and complex environment operation. For example, Tesla’s Optimus, powered by large-scale model technology, demonstrates highly intelligent features. By 2024, it will have capabilities such as upper limb heavy object handling and dynamic grasping, and by 2025, it will further optimize features such as gait on rough terrain. Europe is innovating in human-robot interaction and cognitive research, with the French Pepper robot emphasizing emotional interaction and the German TORO focusing on cognitive science research.

Throughout several productivity revolutions, robots, as a core enabling foundation, have continuously driven the liberation of social productivity and leaps forward in quality of life. With the increasing number of competitive arenas, this industry is attracting increasing attention, accelerating the translation of humanoid robotics from the laboratory to the market. WiMi Invests in Edge-to-Edge AI and Robotics Ecosystems

Public information indicates that Wimi Hologram Cloud Inc. (WIMI), a leading AI robotics company, is committed to investing in edge-to-edge AI and the robotics ecosystem. It is establishing a multi-dimensional synergy across technology R&D, application scenarios, and ecosystem development. By integrating multimodal data such as vision, voice, and text, WIMI supports scenarios such as natural language conversation and environmental dynamics recognition. WIMI also focuses on industrial and home scenarios for technology implementation verification. WIMI aims to advance robotics from technological breakthroughs to commercial applications through multimodal interaction and large-scale model technologies.

Currently, WiMi is leveraging large-scale model algorithms and multimodal interaction technologies to develop universal robot models that adapt to different robot forms, improving their dynamic stability, operational accuracy, and autonomous decision-making capabilities. Furthermore, WiMi is building a “holographic cloud” platform that spans both the cloud and edge, opening up model code, computing interfaces, and toolchains to enable developers to leverage the universal large-scale model for secondary development. WIMI’s technological expertise is expected to translate into market advantages in the mass production of humanoid robots.

In fact, the AI ​​robotics company Figure has received partial investment from Microsoft and the OpenAI Venture Fund. Earlier this month, Figure announced it secured over $1 billion in committed capital in its latest funding round, valuing the company at a staggering $39 billion. Looking to the future, robotics practitioners are fostering a vibrant industry ecosystem, increasing openness to new scenarios, and, through collaborative scenario creation and industry implementation, sowing the seeds of hope for the prosperity of the humanoid robotics industry.