AutoLife Robotics 智动未来Book a demo
Next-generation embodied-intelligence robot

AutoLife S3

An integrated embodied-intelligence platform for perception, mobility and manipulation

AutoLife S3 integrates multimodal perception, autonomous navigation, whole-body motion control and object manipulation in one robot platform, supporting a complete path from environmental understanding to task execution.

AutoLife S3 front product view
  • High-resolution vision and spatial perception

    Stereo AI depth vision supports up to 2K@60Hz transmission for object recognition, spatial understanding and remote task monitoring.

  • Manipulation skills trained through imitation learning

    Demonstrations and physics simulation train grasping and pick-and-place skills to adapt across related objects, motions and task variations.

  • Fine manipulation across common objects

    Seven degrees of freedom per arm + six degrees of freedom in the dexterous hand, together with visual positioning, support identification, grasping, movement and precise placement.

  • Multi-sensor localization and autonomous navigation

    LiDAR, IMU and vision combine for indoor localization, mapping, path planning and dynamic obstacle avoidance within assessed environments.

  • Local computing for embodied-intelligence tasks

    686 AI TOPS of compute supports perception, localization, navigation, motion control and task execution on the robot.

System-level engineering for real environments

Omnidirectional mobility

50 × 34.6 cm Mecanum-wheel base

A compact base supports forward, lateral and in-place movement around assessed aisles, counters and shelves.

Whole-body motion

36-DoF coordinated control

Coordinated neck, waist, arm and base control adapts observation, manipulation and handover postures across task heights.

Manipulation intelligence

Skills trained through imitation learning and physics simulation

Real demonstrations and physics simulation train manipulation policies for target objects, tools and operating conditions.

Body architecture

Compact height-adjustable body

Height adjusts from 106–168 cm, balancing reach with the constraints of stores, showrooms and restaurants.

Multimodal perception

Vision, LiDAR and inertial fusion

Stereo depth vision, LiDAR and IMU data support mapping, localization, path planning and obstacle detection.

AutoLife S3 technical specifications

Computing and energy

Processor
Up to 5.1 GHz / 22 cores
AI compute
686 AI TOPS
Rated runtime
Over 10 hours with autonomous charging

Body architecture and motion

Whole-body degrees of freedom
36 DoF
Rated payload per arm
3–5 kg
Arm degrees of freedom
7 DoF per arm + 6 DoF dexterous hand
Neck degrees of freedom
Up to 3-DoF biomimetic neck

Multimodal perception and autonomous navigation

End-effector vision
HD cameras in both hands
Mobility and localization
Mecanum-wheel base + 3D LiDAR SLAM + IMU
Primary vision
2K@60Hz stereo AI depth vision
3D visual transmission
Up to 2K@60Hz

Dimensions and weight

Robot height
Adjustable from 106–168 cm
Base dimensions
50 × 34.6 cm
Robot weight
Approx. 75 kg