Physical AI Blueprint family

Overview

Generative AI developed in the 2020s possesses the ability to "create" content, enabling anyone to utilize AI through natural language. Furthermore, recent generative AI has begun to exhibit logical thinking and reasoning capabilities, advancing to a stage where it simulates more abstract thought processes. AI agents that understand objectives, formulate plans, autonomously master tools, and execute tasks are now emerging. Beyond this, the emergence of Physical AI—a fusion of general-purpose artificial intelligence like AI agents and robotics—is anticipated, heralding a paradigm shift for humanity. The most significant change will be AI acquiring a physical body through robots and sensors, enabling it to autonomously accumulate "experience" in the physical world and gain new knowledge. This is expected to lead to the emergence of general-purpose robots capable of understanding human verbal instructions and acting autonomously based on their own judgments, even in unfamiliar environments. Such robots hold the potential for AI to become a partner for humans not only in cyberspace but also in physical space, significantly transforming our lives.

Within this Blueprint Family, we will release software stacks for realizing sensor networks for robots and edge AI, essential for achieving Physical AI.

Open

Family Template

Blueprints in this Family

Proposal Presentation