AGIBOT A3 Ultra | Humanoid Robot | 51DOF

The AGIBOT A3 Ultra Developer is aimed at companies, research institutions and development teams wishing to actively advance the field of humanoid robotics. Unlike the standard version, the Developer model offers an open platform for secondary development, custom software, sensor integration and bespoke applications. This allows existing functions to be specifically expanded and adapted to specific project requirements. The platform is suitable for AI research, autonomous systems, pilot projects and the development of new human-robot interaction concepts.

The A3 Ultra Developer features a comprehensive sensor system for navigation, interaction and autonomous operations. A Mid360 LiDAR, an Intel RealSense D457, an Intel RealSense D415, and additional RGB and RGB-D sensing enable omnidirectional perception of the environment. GPS, GPS-RTK and UWB complement the positioning capabilities. Sensor fusion provides the basis for localisation, navigation, visual servoing and perception-based walking on a variety of surfaces. This provides developers with a comprehensive perception platform for demanding robotics applications.

With 51 active degrees of freedom, a peak joint torque of 320 Nm and a speed of up to 1.8 m/s when walking or 5 m/s when running, the A3 Ultra Developer combines dynamic movement capabilities with autonomous mobility. The platform supports localisation, navigation, all-terrain perceptual walking, teleoperation beyond line of sight, and autonomous charging. This enables developers to create mobile applications on a humanoid platform that actively perceives its environment and reacts independently to changes.

The A3 Ultra Developer is supplied with OmniHand-10 hands, which have ten degrees of freedom and support precise manipulation tasks. In conjunction with visual servoing, movements can be executed on the basis of visual information. An 8-channel microphone array, two 5 W loudspeakers and a curved widescreen display are available for interaction. The combination of perception, manipulation, speech, audio output and display creates a versatile development platform for demanding human-robot interaction and automation projects.