Fourier GR-1 - Humanoid With an Emotional Touch

The GR-1 Breaks the Language Barrier

The Fourier GR-1 humanoid represents a leap forward in robotics by blending robotics and AI with human interaction. This robot is designed to communicate with you like a human might. Instead of relying on complex programming, the GR-1 uses advanced AI and a large language model (LLM) to understand regular, everyday language. This makes interacting with it as easy as talking to a person!

What’s cool is that the GR-1 doesn’t just follow basic commands; it can adapt on the fly thanks to its logical reasoning skills. So, if your needs or the situation changes, the GR-1 is smart enough to adjust and keep up.

To make it even more lifelike, the GR-1 is fitted with a curved high-resolution display, powered by AI, that simulates emotions. It also has a real-sense depth camera as its eyes, which helps it perceive the world around it. Add in a microphone and speaker for natural communication, and this robot feels more like a companion than just a piece of tech!

Talk like a Human, Walk like a Human!

The GR-1 isn’t just designed just to stand around and talk. With 2 degrees of freedom (DOF) in its ankle joints, this robot can walk with straightened legs, much like a human. Plus, its self-adapting algorithms enable it to move through uneven surfaces like gravel and grass. It can even handle moving up and down hills!

What’s impressive is how it stays stable. Whether standing still or walking, the GR-1 can handle external forces and self-correct to avoid falling. So, not only does it move smoothly, but it also has the balance to keep going, even in challenging environments.

Mimicking the Human Form & Packed with Power

The mechanical structure of the GR-1 humanoid is designed to mimic human movement. It has a total of 3 degrees of freedom in its head, 3 in its waist, 7 in each arm and 6 in each leg. This provides a high level of dexterity and flexibility, which allows the robot to balance and perform complex tasks. It is powered by 32 actuators with peak performance, each integrated with a motor, driver, reducer, and encoder. These actuators provide a peak torque output of 230 Newton meters, enabling the robot to perform a variety of tasks with strength and precision.

Seeing is Believing: Advanced Perception and Obstacle Avoidance

The GR-1 has an exceptional ability in terms of visual perception and interaction with its environment. Through the use of depth cameras and AI algorithms, the robot can perceive its surroundings and autonomously avoid obstacles and collisions. Its comprehensive perception also allows it to perform reliable object manipulation, such as gripping and handling objects.

Open Source for Open Minds

The platform open sources the robot’s app code, a basic LLM and software development kits to allow for greater collaboration and innovation from people across the globe. It also comes with an app that supports multi-device control, which offers a user-friendly way to manage its functions. The GR-1’s versatility opens itself up to a range of applications in research, education, entertainment, manufacturing and medical rehabilitation.