Consists of standard common modules: Robot Control Unit(RCU), Central Computing Unit (CCU), Sensors&Control and our smart joints SCA.
Follows Cloudminds' Unified Hardware Architecture
Unified Architecture of Cloud robot devices consists of six major systems:
computational system, perception system, communication system, control system, execution system and interactive system.
Powerfull Computing System
*"Cloud-Edge-Device“ hetergenous distributed computing platform built with distributed processing units of CPUs and GPUs.
* Provides powerful local and edge computing power as well as AI capabilities for robot auto navigation, image processing, motion control, networking communicaiton and parallel processing of sensor data.
Robotic Perception System
*Multiple multi-modal perception sensors such as cameras, microphone arrays and LiDAR .
*Provide precise perception of cloud robot running in the real physical world.
Secure Network System
By using a variety of technologies such as our proprietary PonCAN high speed optical communication, CANFD and Ethernet, the system of the internal robot can be connected at high speed, efficiently and in real time.
Control System
* Based on AI algorithms such as path planning, motion control, and motion balance.
* Control robot movement and navigation, obstacle avoidance, joint movement and anthropomorphic movements of the robot body.
Execution System
* Built with multiple smart joints (Smart Compliant Actuator) and local execution processors .
* Execution of motion and action instruction, including hand waving, hand shaking, pushing, pulling, rowing, and AI guided grabbing.
Interactive System
The robot controller, as an agent of the "cloud brain", collects information perceived by the multi sensors perception system in real time, issues control instructions to each system on robot device by collaborating with the "cloud brain". Meanwhile, users can interact with robots through multi-model interactive interfaces such voice recognition , visual detection, touch screen via robot controller.
