Skip to main content
Cyberwave supports a range of robotic hardware from articulated arms to mobile rovers, quadrupeds, and standalone cameras. Each platform integrates with Cyberwave through a digital twin, giving you real-time control, telemetry streaming, dataset recording, and AI model deployment from a single interface.

SO101 Robot Arms

The SO101 is an open-source, 6-degree-of-freedom (6-DOF) robotic arm set designed for desk-based manipulation. Built from 3D-printed parts and standard servo motors, it offers a low-cost entry point into real robotic hardware.

Physical Components

  • 6-DOF articulated arm: compact manipulator for close-range, desk-based operations
  • Servo-driven joints: position-controlled motors for real-time joint movement
  • Gripper end-effector: suitable for basic pick-and-place tasks
  • Leader-follower configuration (optional): dual-arm setup where the leader captures human motion and the follower mirrors it in real time
  • USB control: direct connection from a laptop or SBC (Raspberry Pi) over USB/serial

What You Can Do with Cyberwave

  • Onboard from the catalog and create a digital twin automatically
  • Teleoperate using a physical leader arm with joint-level mirroring
  • Remote operate without a leader arm and control directly from the dashboard, SDK, or API
  • Assign controller policies (keyboard, gamepad, VLA models)
  • Record and export episodic datasets for training
  • Train ML models and deploy them as autonomous controller policies
  • Test in simulation, then deploy to the physical arm without code changes

SO101 Setup Guide

Connect, calibrate, and start using the SO101 with Cyberwave

UGV Beast Rover

The UGV Beast is an off-road tracked open-source AI robot with a dual-controller architecture: a Raspberry Pi for high-level tasks (planning, perception, AI) and an ESP32 microcontroller for low-level motor control (motors, encoders, IMU, LEDs).

Architecture

The system operates in three layers:
  • Cyberwave Cloud: hosts the digital twin, dashboard, MQTT broker, and backend APIs for monitoring, commanding, and training
  • Cyberwave Edge (Raspberry Pi): runs the ROS 2 stack and MQTT bridge inside a Docker container, translating between ROS 2 topics and Cyberwave MQTT
  • Robotic Hardware (ESP32): handles low-level motor control, sensors, camera pan-tilt, and LEDs over serial UART

What You Can Do with Cyberwave

  • Add a UGV Beast from the catalog and start streaming telemetry immediately
  • Control the rover in real time via keyboard from the dashboard
  • Stream the onboard camera feed to the digital twin
  • Record driving sessions (camera, IMU, controls) into episodic datasets
  • Train navigation models and deploy them as autonomous policies
  • Test in simulation, then deploy to the physical rover

UGV Beast Setup Guide

Install the Cyberwave Docker image and launch the stack on your rover

Unitree Go2 Quadruped

The Unitree Go2 is an intelligent quadruped robot designed for research, exploration, and real-world applications. It combines AI-driven locomotion with robust mechanical performance.

Key Specifications

SpecDetail
Recognition SystemUltra-wide 4D LiDAR L1 (360° x 90° FOV)
Max Speed~5 m/s
Peak Joint Torque~45 N·m
WirelessWi-Fi 6, Bluetooth, 4G
Battery8,000 mAh (standard) / 15,000 mAh (extended)
Endurance~2–4 hours depending on configuration

Core Capabilities

  • 4D LiDAR perception: hemispherical FOV with 0.05m minimum detection for all-terrain navigation
  • AI-powered locomotion: trained via large-scale simulation for obstacle climbing, recovery, and adaptive gaits
  • 3D mapping and autonomy: point cloud mapping and path planning using built-in LiDAR
  • Intelligent side-follow (ISS 2.0): wireless vector positioning with obstacle avoidance
  • OTA updates: continuous firmware improvements via the cloud
For detailed hardware specifications and assembly instructions, refer to the official Unitree Go2 documentation.

Go2 Setup Guide

Connect an edge device to the Go2 and pair it with Cyberwave

Camera

A camera twin is a digital representation of a physical camera within your Cyberwave environment. It streams live video into the platform, enabling real-time monitoring, data recording, and AI-powered vision workflows.

Supported Camera Types

  • USB webcams (Logitech, generic UVC-compatible cameras)
  • Laptop built-in webcams
Once connected, your camera feed is available across the platform in the environment viewer, for dataset recording, and as input to ML models and workflows.

Camera Setup Guide

Connect a webcam or laptop camera to Cyberwave

Hardware at a Glance

PlatformTypeConnectionUse Cases
SO1016-DOF robot armUSB/serialManipulation, teleoperation, imitation learning
UGV BeastTracked roverWi-Fi (MQTT)Navigation, mapping, autonomous driving
Unitree Go2QuadrupedEthernet to edge deviceLocomotion, patrol, inspection
CameraVisual sensorUSBMonitoring, vision workflows, dataset recording