Donut Robotics Unveils Cinnamon 1, the First Mass-Produced Gesture-Controlled Humanoid from a Japanese Brand

A New Gesture Era

In noisy factories and quiet retail spaces where voice commands fail, controlling robots requires a more intuitive approach, and Donut Robotics has answered this challenge with the world premiere of Cinnamon 1. The Tokyo-based startup officially announced its first mass-produced bipedal humanoid robot, a Japanese-branded device designed to replace repetitive work in construction sites and factories by 2026. This launch marks a significant evolution from the company's earlier communication robots, introducing a system that allows operators to guide the machine using simple hand and finger movements without speaking.

Why It Matters

Cinnamon 1 stands out by prioritizing silent gesture control over traditional voice interfaces, enabling natural coordination with human coworkers in environments too loud for shouting. Its core innovation lies in the integration of Vision-Language-Action AI, which interprets visual cues and linguistic intent to execute actions with high fidelity. The robot also features full-body coordination and teleoperation capabilities that allow it to mimic human movements for tasks that are risky or repetitive for people. Cinnamon 1 is less about raw hardware and more about redefining how humans silently control robots in complex workspaces.

Motion to Action Flow

The system operates through a clear input-to-output flow where human gesture input is captured by RGB and stereo cameras before being processed by the onboard VLA AI model. This artificial intelligence analyzes the hand movements and contextual language to determine the necessary joint actuation and balance correction required for the next step. The result is a real-time response where the bipedal robot executes the commanded task while maintaining stability through its visual SLAM and LiDAR mapping navigation systems.

Factory Floor Focus

In a construction site monitoring scenario, a foreman can silently direct Cinnamon 1 to inspect a specific structural element by pointing or using a specific hand signal, avoiding the need to shout over machinery noise. The robot uses its LiDAR and ultrasonic sensors to navigate the uneven terrain safely while carrying tools or precision instruments to the designated location. This capability ensures that critical safety checks and material deliveries happen efficiently without disrupting the workflow of human workers who are operating in hazardous conditions.

Capabilities in Detail

The Cinnamon 1 is engineered to be a robust industrial partner, standing approximately 170 cm tall with a weight of 70 kg that allows it to handle tools and packages effectively. It moves at a speed of 4.5 km/h (1.25 m/s), which is sufficient for rapid response in large facilities, while its battery system is designed for a four-year usable life typical of industrial lithium-ion setups. Equipped with an array of sensors including LiDAR, IMU, and force sensors, the robot ensures safe collaborative mode operation with force limiting and collision detection to protect human operators.

Rivals Edge Check

Industry Direction

This development signals a broader industry shift toward teleoperation-first humanoids that rely on human guidance rather than pre-programmed motion libraries for autonomy. Manufacturers are increasingly recognizing that full autonomy in dynamic, unstructured environments remains a challenge, making intuitive control systems like Silent Gesture Control a critical differentiator. The move suggests that the next wave of humanoid adoption will prioritize robots that can seamlessly integrate into existing human workflows through non-verbal communication rather than replacing human oversight entirely.